Signals
Community chatter from Hacker News, Bluesky, Mastodon, Lobsters, Lemmy & GitHub issues that mentions a tracked vendor — unverified early warnings, not confirmed changes. Catch a price, limit or deprecation move before it's official, then confirm it at the source. Verified changes live on the change feed →
Unverified · community-sourced · confirm before acting
-
FYI: OpenAI text watermark detection drops to 17% after 25% of words change: About 80% of 200-token passages are caught at a 1% false-positive target. API users can opt in now; ChatGPT and Codex text in the EU is marked within weeks. https:// ppc.land/openai-text-watermark -detection-drops-to-17-after-25-of-words-change/ # OpenAI # AI # MachineLearning # TextWatermarking # ChatGPT
-
💸 DeepSeek V4 Pro 0813 input price dropped 48% — $0.300 → $0.158/MTok. Compare providers, speed & capability. #LLM #AI
-
💸 DeepSeek V4 Flash 0423 input price dropped 65% — $0.017 → $0.0057/MTok. Compare providers, speed & capability. #LLM #AI
-
@DailyCyberSecurity@infosec.exchange on Mastodon
OpenAI introduces the GPT-6.1 Sol Ultrafast version for API, Codex, and ChatGPT Work. Experience AI speeds up to eight times faster than the standard edition. # OpenAI # GPT6 # ArtificialIntelligence # ChatGPTWork # Codex https:// dailytechnow.com/openai-gpt-6- 1-sol-ultrafast/
-
OpenAI API pricing changed OpenAI pricing page now displays Business and Enterprise ChatGPT plans with per-seat subscription model ($20-$100/month) https://changeradar.ai/tools/openai
-
Omit deprecated Gemini sampling parameters for newer models
groupthinking/EventRelay — Prevent Gemini 3+ requests from sending deprecated sampling overrides (e.g., temperature) while preserving overrides for older models and other providers.
-
@feed@igeek.gamer-geek-news.com on Mastodon
🤖 LegalOn halves Codex costs while maintaining development speed LegalOn cut estimated daily Codex costs by 65% while maintaining development speed. It matched Astra, Sol, and Luna to tasks and managed budgets strategically. 📰 Source: OpenAI News 🔗 Link: https://openai.com/index/legalon-halves-codex-costs # AI # ArtificialIntelligence
-
Anthropic LLM provider fails on current Claude models: "temperature is deprecated for this model"
OneUptime/oneuptime — ### Anthropic LLM provider fails on current Claude models: "temperature is deprecated for this model" Describe the bug The Anthropic path in LLMService.getAnthropicCompletion (packages/Common/Server/Utils/LLM/LLMService.ts) always sends temperature (request.temperature?? 0.7; incident/alert features pass 0.2, the provider test passes 0). Current Claude models reject any value other than 1 with HTTP 400: {"type":"error","error":{"type":"invalid_request_error","message":"temp
-
@claudecodechanges.bsky.social on Bluesky
[27/36] Claude apps gateway now attempts non-listed Bedrock models in your region instead of refusing. Headless rate limit warnings now indicate extra usage status. Plugin-authoring skill avoids terminal commands for non-terminal sessions. Grep now handles -l, -c, -r flags correctly.
-
[Codex] GPT-6.1 Sol Ultrafast Mode API Pricing OpenAI announced GPT-6.1 Sol Ultrafast mode API pricing at $12 per million input tokens and $60 … Source: https://x.com/OpenAIDevs/status/2108262830348886150 Details: https://quotatick.com/en/events/27ba85b3-b652-439b-aa4e-f638270b290b #Codex #AIcoding
-
@researchbuzz_firehose@rbfirehose.com on Mastodon
The Register: Anthropic Claude subscription plan provides more value than OpenAI’s, study says. “Following changes to OpenAI’s $200 per month subscription plan that effectively halved its value, SemiAnalysis looked at the various subscription plans offered by AI companies, because that’s still the main way that individuals and small businesses buy AI services.” https://rbfirehose.com/2026/10/08/the-register-anthropic-claude-subscription-plan-provides-more-value-than-openais-study-says/
-
@researchbuzz_firehose@rbfirehose.com on Mastodon
Gizmodo: OpenAI Decides ChatGPT Would Be Better With More Sponsored Results. “Hey! You got ads in my AI responses! On Monday, OpenAI announced that it plans to start expanding the types of advertisements that appear in ChatGPT alongside the model’s outputs.” https://rbfirehose.com/2026/10/08/gizmodo-openai-decides-chatgpt-would-be-better-with-more-sponsored-results/
-
@hackaday@www.urbanmind.net on Mastodon
OpenAI is rolling out Ultrafast mode for GPT-6.1 Sol across its API, Codex and ChatGPT Work. The... # ai # automation # tools # openai # software # coding # development # engineering # inclusive # community GPT-6.1 Sol Ultrafast Rolls Out Across the API, Codex and ChatGPT Work
-
Switch the solver to free models (Gemini 3.8 Flash free tier + one free fallback)
AbhishekBalija/MathGPT — ## What Move the solver to free models: Primary: Gemini 3.8 Flash on the Gemini API free tier (same key, GEMINI_MATH_AI_API). Fallback: one free OpenRouter model, picked by the test below. Candidates seen on 2026-10-09: nvidia/nemotron-3-super-120b-a12b:free (structured output), google/gemma-4-31b-it:free (images). Why Both current backups (google/gemini-2.0-flash-001, tngtech/deepseek-r1t2-chimera:free) no longer exist on OpenRouter, and gemini-2.5-flash is legacy. USE_
-
Ultrafast is rolling out today for GPT-6.1 Sol in the API, Codex, and ChatGPT Work. Near-Astra intelligence at up to 8x faster speeds than Sol Standard, so you can build as fast as the ideas come. https://t.co/Cyy9k2hifV
-
fix(engine): read the reset time of a Claude Code usage limit correctly
Mobius-Toolkit/Mobius — Goal Make the pause of a Claude Code usage limit end at the correct reset time. Cause (see the result in #347) decodeObject (internal/engine/transcript.go) calls decoder.UseNumber(), so each JSON number is a json.Number. Agent.record (internal/engine/agents.go) reads _meta._claude/rateLimit.resetsAt with.(float64). This type assertion always fails, so resetHint stays zero. Then detect uses resetsAt (internal/engine/limits.go). It reads "resets 5:10pm (Europe/Warsaw)", ign
-
@marcus-schuler.com on Bluesky
Google Playground ships Wednesday for US adults 18+. Text-to-game conversion via Gemini and Nano. Weekly tokens for creation; browsing free. Unity Spark (advanced tier) in beta later. https://www.implicator.ai/google-launches-playground-letting-us-adults-build-browser-games-from-ai-prompts/
-
@DailyCyberSecurity@infosec.exchange on Mastodon
OpenAI text watermarks are launching in the European Union for ChatGPT and Codex outputs. The textGrain technology ensures transparency law compliance. # OpenAI # ChatGPT # TextWatermarks # EuropeanUnion # TextGrain https:// dailytechnow.com/openai-text-w atermarks/
-
3 ways to ask an AI a yes/no question right now: openai decisions API (public beta, gpt-6-luna): $0.10 per 1M input tokens, reads images too jev: $0.042, text only strands decider 2B: open source, $0, ~115ms median on an RTX 3090 no output tokens. you get a probability back, not an essay. the most useful boring AI category https:// developers.openai.com/api/docs /guides/decisions # AI # LLM # OpenSource # DevTools
-
TabularisDB/tabularis — ### Describe the bug With the Anthropic provider, AI Assist / Generate SQL fails whenever the default model is a Claude 5.5 model (claude-sonnet-5-5, claude-haiku-5-5). The Anthropic API rejects the request: Anthropic Error: {"type":"error","error":{"type":"invalid_request_error","message":"temperature is deprecated for this model."},"request_id":"req_011Cfq7zz6YeCsc1BR8UZJpx"} The same setup works with Claude 4.7 models, so the API key and configuration are fine. The cau
-
@thoughtsbrewingllc.bsky.social on Bluesky
Gemini Model Access Changes: What You May Lose Watch Here -> https://www.youtube.com/watch?v=kdkg1Fiups0
-
Gemini 3: thinkingBudget: 0 sent when reasoning is disabled; deprecated, will 400 on upcoming models
NousResearch/hermes-agent — ## Summary _build_gemini_thinking_config (agent/transports/chat_completions.py) sends thinkingBudget: 0 for gemini-3* and gemini-flash-latest when reasoning is disabled. Google has announced that thinking_budget is deprecated for Gemini 3: requests that set it will return 400 INVALID_ARGUMENT on upcoming models (today it is still silently remapped to thinking_level). Google AI Studio is already emailing users ("[Action Required] Update thinking_budget and sampling par
-
@mobileff.bsky.social on Bluesky
おっと、またClaude CodeがWeekly Limitいってしまった・・・ 明日いっぱい動かせないのはちょっとつらいなー
-
MervinPraison/PraisonAIDocs — ## Context Upstream PR MervinPraison/PraisonAI#5344 (merged on 2026-10-08) changes the default model for the Python realtime capability and its CLI from the OpenAI-deprecated gpt-4o-realtime-preview to the GA (non-deprecated) gpt-realtime. The SDK is now the source of truth for gpt-realtime; existing docs still show the deprecated id and must be brought back in sync. Upstream PR: https://github.com/MervinPraison/PraisonAI/pull/5344 Upstream issue: https://github.com
-
Anthropic rate limits changed Updated Rate limits page URL from /settings/limits to /usage/limits; changed 'Request rate limit increase' button label to 'Request tier increase'. https://changeradar.ai/tools/claude-code
-
OpenAI's GPT-6 Luna rolled out to free ChatGPT users Thursday with lower safety scores on teen-focused tests than its predecessor. Sexual-content safety fell to 0.904 from 0.984. The company says stricter teen provisions still apply, but the regression raises questions about tradeoffs in new model releases. https://www. implicator.ai/openais-gpt-6-re aches-free-chatgpt-users-after-scoring-lower-on-teen-safety-tests/ # AI # SafetyTesting # OpenAI
-
@researchbuzz_firehose@rbfirehose.com on Mastodon
Politico: German search engine ditches Mistral, bets on Chinese open-source AI. “A Berlin-based search engine used by government services is dropping its French AI partner, Mistral, in favor of open-source models, including those from China, its CEO told POLITICO.” https://rbfirehose.com/2026/10/08/politico-german-search-engine-ditches-mistral-bets-on-chinese-open-source-ai/
-
OpenAI has officially begun the global rollout of GPT-6, introducing an Intelligent UI feature to ChatGPT. Source: Innovation Village https:// innovation-village.com/openai- rolls-out-gpt-6-and-intelligent-ui-to-all-chatgpt-tiers/ # AI # OpenAI # ChatGPT
-
https:// winbuzzer.com/2026/10/08/claud e-haiku-5-5-cuts-ai-token-prices-up-to-90-xcxwbn/ Anthropic's new Claude Haiku 5.5 small AI model adds adjustable reasoning for summaries and delegated tasks and beats OpenAI's cheap GPT-6 Luna model in benchmarks. # AI # Anthropic # OpenAI # Claude # Haiku55 # GPT6Luna
-
OpenAI provider sends deprecated max_tokens, which newer models reject
nucket/NekoAI — ## Summary OpenAIProvider sends max_tokens. Newer OpenAI models (o-series, gpt-5 family) reject this parameter with HTTP 400 and expect max_completion_tokens. A user who types one of those models into Settings gets "Sorry, something went wrong". Where src/ai/providers/openai.ts (line 23) Proposed fix Send max_completion_tokens (accepted by current chat models). Consider updating the default model (gpt-4o-mini) as well. Acceptance criteria [ ] Chat works with both gpt-4o-mini and
-
repo-ci-baseline: Claude Review on Opus can strand a PR when the subscription's usage limit is hit
mpaulosky/dotfiles — ## Problem Statement Claude Review authenticates with CLAUDE_CODE_OAUTH_TOKEN, a subscription token from claude setup-token, not an API key. Since #120 it runs Opus 5.5 at high effort, which draws on the plan's Opus usage limits. ADR 0004 priced it as about $1.20 a review, which is API-style billing. In a re-Apply round (six repos, a review on every new head while review:claude is on, several minutes of Opus each), the account can hit its usage limit. A refused run fails the
-
beeltec/aiproxy — ## Problem aiproxy does not expose the native OpenAI Decisions API. Model-only pricing cannot distinguish Decisions billing from normal Luna requests. Scope Add native POST /v1/decisions for OpenAI API-key connections and gpt-6-luna. Preserve native payloads and answers, model routing, permissions, limits, cancellation, and usage accounting. Add endpoint-aware price selection, versions, overrides, long-context rates, and regional billing facts. Update endpoint and pricing docum
-
📰 Latent — Haiku 5.5 at $0.10/M, plus Mistral's Le Chonk • Claude Haiku 5.5: 1M context at $0.10 per million input tokens • Mistral says open-weight 'Le Chonk' can rival top closed models • ChatGPT's 'Intelligent UI' answers with charts, forms and buttons • Perplexity releases pplx-embed-v2-late, MIT-licensed, 0.6B and 9B • Microsoft gives Copilot access to local files and OS-level actions Today's AI briefing → https:// latent.steenbergen.ai/archive/ 2026-10-08/ # AI # MachineLearning # TechNew
-
Anthropic、Claude Haiku 5.5 を発表、スモールモデル API の価格を引き下げ – Unite.AI https://www. yayafa.com/2905481/ # AgenticAi # AGI # AI # ArtificialGeneralIntelligence # ArtificialIntelligence # エージェント型AI # 人工知能 # 汎用人工知能
-
💸 DeepSeek V4.1 Flash input price dropped 63% — $0.050 → $0.018/MTok. Compare providers, speed & capability. #LLM #AI
-
KempnerInstitute/hpc-agentic-recipes — ### Describe the Feature Claude Code now marks ANTHROPIC_SMALL_FAST_MODEL as deprecated and uses ANTHROPIC_DEFAULT_HAIKU_MODEL instead (see the environment variables reference ). The README, docs/quickstart.md, the recipe README template, and 27 recipe README and client.env files still set the old name. Switch them to ANTHROPIC_DEFAULT_HAIKU_MODEL. Use Case Users who copy a recipe's client settings get the supported variable, and the computing handbook can
-
aiur-team/aiur — ## Follow-up to #3029 / #3038 When an account hits its usage limit, the session handoff (#3038) moves only claude-repl sessions to another account. Headless claude workers, which run through the aiur-claude app-server, stay on wait-for-reset. The app-server keeps its thread map in memory, and thread/start cannot seed a moved session transcript (see handoff_candidate? in src/lib/aiur/orchestrator/rate_limit_fallback.ex). Done when Resume support. aiur-claude can resume a session
-
Add retry with exponential backoff for Gemini API free tier
andreshg112/etoro-trading — Currently, calls to the Gemini API don't retry on 429/503 errors. Adding an exponential backoff wrapper would make the bot much more reliable when running on Google AI Studio's free tier.
-
Zenvi-pro/zenvi-core —
-
GitHub has made Claude Haiku 5.5 generally available in Copilot for quick coding edits, terminal tasks and subagents, across Pro, Pro+, Max, Business and Enterprise plans. Usage is billed at provider list pricing, and access is rolling out gradually.
-
📉 GLM 5.3 Flash (batch) dropped 60% in 30 days: $0.5 to $0.2 per 1M output tokens, with open weights. For batch inference, that pricing shifts the cost math. https:// olud.ai/model/glm-5-3-flash-ba tch.html # AI # LLM # Pricing
-
Claude Haiku 5.5: benchmarks, pricing, what's new https:// velstech.net/claude-haiku-5-5. html?utm_source=mastodon&utm_medium=social&utm_campaign=new-post # llm # anthropic # claude
-
Claude Haiku 5.5: benchmarks, pricing, what's new https:// velstech.net/claude-haiku-5-5. html?utm_source=mastodon&utm_medium=social&utm_campaign=new-post # llm # anthropic # claude
-
@simon_willison@robot.villas on Mastodon
Introducing Claude Haiku 5.5 https://simonwillison.net/2026/Oct/7/claude-haiku-5-5/ # AI # LLM # Haiku
-
Build richer, more interactive plugins for ChatGPT. With plugin extensions, you can launch plugins from the sidebar, support @ mentions, and add file viewers. See how in this walkthrough with @coreyching: https://t.co/Oqq1entWNT
-
OpenAI is introducing Intelligent UI alongside GPT-6, bringing interactive charts, diagrams and task-specific tools into ChatGPT conversations. These are meant to be used or adjusted, not just viewed: examples include editable graphs and a personal savings calculator. The global rollout starts with Pro, Plus, Business and Enterprise subscribers, followed by free and Go users. OpenAI says the aim is to make complex topics eas… https:// en.hacks.gr/me-to-gpt-6-to-cha tgpt-tha-prosthetei-diagrammat
-
OpenAI launched GPT-6 for ChatGPT, deploying dynamic UI generation to its 1.2 billion weekly users. The system interleaves thinking with answering, cutting search latency by 44%. Paid accounts upgrade today; free users transition tomorrow. Equipping an LLM to compile custom widgets on the fly turns the chat window into a disposable app generator, destroying the unit economics of standalone utility software. # OpenAI # ChatGPT https:// zubiqo.com/news/openai-launche s-gpt-6-with-intelligent-ui-tu
-
@feed@igeek.gamer-geek-news.com on Mastodon
📰 ChatGPT’s ‘Intelligent UI’ update fills its responses with pictures, charts, and buttons OpenAI is launching a new Intelligent UI feature in ChatGPT that allows the chatbot to answer your questions with interactive visuals. The update, which is rolling out to all users alongside GPT-6,... 📰 Source: The Verge 🔗 Link: https://www.theverge.com/ai-artificial-intelligence/1007276/openai-chatgpt-intelligent-ui-gpt-6 # AI # ArtificialIntelligence # ChatGPT
-
@AndriiKuznietsov75@social.kyiv.dcomm.net.ua on Mastodon
🤖🫰Аналітики SemiAnalysis протестували ліміти усіх популярних ШІ-підписок: # Anthropic, # OpenAI, # Meta, # SpaceXAI, # MiniMax, # Moonshot, Z.ai, # CursorCognition. Головний висновок такий: на моделях середнього рівня # Claude дає приблизно в 4–5 разів більше API-еквівалентної вартості, ніж OpenAI. Після останніх урізань у # ChatGPT розрив став ще помітнішим. https:// highload.tech/uk/claude-proty- chatgpt-eksperty-pereviryly-yaki-pidpysky-najbilsh-vygidni/
-
@sneptech.bsky.social on Bluesky
i have no idea because for some godawful reason claude desktop doesn't say, but i know it's used about 15% of a Max X20 weekly limit
-
Claude Haiku 5.5 is now available in Cursor! On shorter requests, it costs 10x less than Claude Haiku 4.5. Turn it on in Cursor Settings > Models. https://t.co/dkYOvcxsLs
-
GPT-6 and Intelligent UI, now rolling out in ChatGPT for everyone. Intelligent UI in ChatGPT delivers fast, interactive answers that make everyday questions more visual, complex topics easier to grasp, and interactive tools for your task available on the spot. https://t.co/XL2gDCPBwG
-
Anthropic rate limits changed Added Claude Haiku 5.5 model to Start tier rate limits (1,000 RPM / 2M ITPM / 400K OTPM); marked Sonnet 4.x as deprecated. https://changeradar.ai/tools/claude-code
-
Introducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released. On average, it costs around 75% less to run than Claude Haiku 4.5. https://t.co/jm06cZJfkV
-
@techcrunch@robot.villas on Mastodon
ChatGPT is getting a lot more visual, with the launch of a new interface https://techcrunch.com/2026/10/07/chatgpt-is-getting-a-lot-more-visual-with-the-launch-of-a-new-interface/ # AI # Tech # OpenAI
-
@paulocastellano.bsky.social on Bluesky
6/6 Official pricing: https://ai.google.dev/gemini-api/docs/pricing Follow me for new AI releases and practical ways to use them in your products.
-
Let users redeem Claude usage limit resets (experimental)
rock3r/headroom — Headroom shows Claude's usage limit resets, but cannot use them yet. Codex and Grok resets can already be used from the reset sheet. What Claude's API needs (from research, not tried live) Redeem with POST https://api.anthropic.com/api/organizations/{org}/reset_rate_limits, with the same OAuth token and anthropic-beta: oauth-2025-04-20. Body: {"program":"cedar_ember","grant_id":" ","request_id":" "}. Only the grant named next_grant_id can be used. The org id is organization.uui
-
@pondero-ai.bsky.social on Bluesky
Full breakdown 👇 https://pondero.ai/news/2026-10-07-gemini-free-tier-flash-lite/?utm_source=bluesky&utm_medium=social&utm_campaign=daily-3df1c59ed79550bd7c95d7e5e37e2555 #AINews #TechNews #LLM
-
yesterday: reflection's beam. 501B params, weights "later this month" today: mistral large 4. 1 trillion params, 49B active, weights "by the end of the month" october is the month of open models you can't download yet lol still, a 1T open model trained in europe is a big deal https:// mistral.ai/news/mistral-large-4 # AI # OpenSource # LLM # Mistral
-
@hackaday@www.urbanmind.net on Mastodon
Counting Claude tokens with tiktoken undercounted my prompts by 14%. 37 of 418 requests failed. Here's the audit and the count_tokens fix. # claude # llm # python # ai # software # coding # development # engineering # inclusive # community tiktoken Undercounts Claude Tokens: 37 of 418 Requests Blew the Limit
-
@feed@igeek.gamer-geek-news.com on Mastodon
🤖 Helping teens learn, plan, and shape the future of AI College Planner is coming to ChatGPT for Teens to help students manage college applications, alongside new flashcards, quizzes, and a teen AI council. 📰 Source: OpenAI News 🔗 Link: https://openai.com/index/teens-learn-and-plan # AI # ArtificialIntelligence # ChatGPT
-
@feed@igeek.gamer-geek-news.com on Mastodon
📰 ChatGPT is getting college planning tools OpenAI is bringing new tools to ChatGPT for Teens, a mode for teens introduced in August with safeguards and break reminders, to help users with the college application process. "College Planner br... 📰 Source: The Verge 🔗 Link: https://www.theverge.com/ai-artificial-intelligence/1005194/openai-chatgpt-teens-college-planner-notecards # AI # ArtificialIntelligence # ChatGPT
-
@verge@www.urbanmind.net on Mastodon
OpenAI is bringing new tools to ChatGPT for Teens, a mode for teens introduced in August with safeguards and break reminders, to help users with the college application process. "College Planner brings together application requirements, deadlines, tasks, and financial-aid steps for schools on a student's list in one plan," OpenAI says. "Students applying to several […] # AI # News # OpenAI ChatGPT is getting college planning tools
-
OpenAI launched a Decisions API for gpt-6-luna that forces deterministic yes/no or categorical outputs instead of standard text. It runs 10x faster at $0.10 per million input tokens, with outputs entirely free. Offering hyper-cheap, zero-retention categorization directly suffocates the enterprise business case for standalone decision models before they can gain traction. # OpenAI # DeveloperTools https:// zubiqo.com/news/openai-launche s-10x-faster-decisions-api-with-free-output-tokens-ew85li
-
Remove deprecated Gemini generation parameters before the next model change
cinatra-ai/gemini-connector — ## Context Google announced that the Gemini API will stop accepting some generation parameters on its upcoming models: thinking_budget (SDK: thinkingConfig.thinkingBudget) will return 400 INVALID_ARGUMENT instead of being remapped. Use a thinking level (SDK: thinkingConfig.thinkingLevel, values minimal, low, medium, high) or omit it. Supported levels differ per model. Google already documents intermittent 400 errors for the deprecated budget on Gemini 3 series model
-
GPT-6.1 Sol launched at OpenAI DevDay: near GPT-6 Astra quality at $2/$10 per M tokens vs $10/$50, with 1.05M context. A 5x price cut at similar capability changes which agent workloads are worth running.
-
OpenAI shuts down gpt-4, gpt-3.5-turbo, o1 and o3-mini on Oct 23. Is your code still using them? One command: pipx run llm-sunset → every deprecated model ID in your repo, its shutdown date, and the replacement. Free, open source, works in CI. https://t.co/NHpuJLbJYs
-
New: Claude Opus 5.5 is cheaper and faster than Opus 5. For teams running agents in production, the pricing change matters more than the benchmark gains Anthropic cut Opus pricing 20-60% across the board and reported real cod https://t.co/jJ8np4HY2Q #AIagents #ARK
-
fix(registry): audit and models-generate ignore models.dev status: deprecated
laurigates/pal-mcp-server — ## What scripts/audit_model_registry.py and scripts/generate_model_entries.py treat every model in a models.dev provider slice as live. Neither reads the per-model status field, which models.dev sets to "deprecated" for withdrawn models. Why it matters After the OpenCode Go sync (the PR that adds space-bunny), just models-audit-one opencode_go_models.json lists space-bunny-free and grok-4.5 as CANDIDATE ADDITIONS. Both carry "status": "deprecated" in models.dev, and b
-
Grok 4.7 launched at $2 per million input tokens — half the price of GPT-5.6 Sol. Everyone's celebrating the price cut. Almost nobody's reading the second page. 🧵 https://t.co/YU6bzmBWKi
-
WIRED tested OpenAI’s Dots, an assistant that can handle online tasks and send updates. It gathered sofa options and improved its suggestions with clearer preferences, but got the user’s name wrong and failed to complete a subscription cancellation after a website check stopped it. The test reports no data exposure or financial loss. Dots is included in a $100/month subscription, and the results are a reminder that delegated tasks… https:// en.hacks.gr/i-openai-paroysias e-to-dots-voitho-poy-ana
-
🍌 Google dropped Nano Banana 2.1 — 4K image gen/edit, handles 14 reference images, keeps 4 characters / 10 objects consistent. API price cut ~50% to $0.0336/1K images. Rolling out to Gemini app and AI Studio. (THE DECODER) https://t.co/KznseegnUm #Google #AI
-
@ampm-aiops.bsky.social on Bluesky
🐾 [Review] Cursor's Free Tier Limits: The Official Page Stopped Naming Numbers Review | Cursor: free tier, limits and pricing checked by AMPM. Last verified 20 https://ampm-aiops.com/en/guides/cursor-review-2026/?utm_source=bluesky&utm_medium=social&utm_campaign=auto
-
LLM: add a free-tier Gemini key to the provider rotation
VishnujanNarayanan/Job_Application_Bot — ## Goal Add a second, free-tier Gemini key to the LLM rotation, so more parses run at $0 and fewer depend on OpenRouter. Details The key comes from a Google AI Studio project with no billing attached, so it can only ever use the free tier. It needs its own provider name, gemini-free. llm.rotation matches on provider names, so a second entry called gemini would pull the paid key into the rotation. It uses gemini-3.5-flash-lite, because gemini-2.5-flash-lit
-
Google Gemini API pricing changed Gemini 3.0 Flash and Flash-8B image generation prices increased up to 49% for 4K output https://changeradar.ai/tools/google-gemini
-
📰 Latent — Mistral's 1T open-weights model, and OpenAI agents gone feral • Mistral Large 4 previews: 1T parameters, open weights promised • OpenAI agents hit Wikipedia with hacking attempts and traffic floods • OpenAI ships its Decisions API • EmbeddingGemma 2: open multimodal embeddings under Apache 2.0 • OpenAI releases another 722 maths manuscripts from an unreleased model Today's AI briefing → https:// latent.steenbergen.ai/archive/ 2026-10-07/ # AI # MachineLearning # TechNews
-
ICYMI: OpenAI text watermark detection drops to 17% after 25% of words change: About 80% of 200-token passages are caught at a 1% false-positive target. API users can opt in now; ChatGPT and Codex text in the EU is marked within weeks. https:// ppc.land/openai-text-watermark -detection-drops-to-17-after-25-of-words-change/ # OpenAI # TextWatermark # AI # MachineLearning # ChatGPT
-
Fix Production Planner deprecated Workers AI model before Research AI
toki1031/creator-tools — PR #419 introduced @cf/meta/llama-3.1-8b-instruct. Cloudflare's May 2026 deprecation notice says this model was deprecated May 30, 2026. Before adding Research AI: replace Planner model with a current Workers Free-compatible multilingual text model; preserve JSON ProductionBrief behavior; add a safe diagnostic/contract check where practical; run regression tests, Quality Gates and Cloudflare Preview; do not merge without explicit confirmation. Current Cloudflare docs lis
-
@tomkalei@mathe.social on Mastodon
openAI released more than 700 Theoremoids: https:// github.com/openai/math/tree/ma in It does include Hilbert’s 10th problem over QQ right on the first page which I have also been prompting (only half-jokingly): https:// machteburch.social/@tomkalei/1 16906337736015636 https:// mathe.social/@tomkalei/1172302 40907462837 So it’s undecidable… told you so.
-
OpenAI drops another batch of mathematical breakthroughs Ready to read about math? https://www. theverge.com/ai-artificial-int elligence/1005004/openai-math-release-github # TopNews # News # OpenAI # AI # theAdvisoryGrouponMathematicsandArtificialIntelligence
-
OpenAI dropped 722 mathematical manuscripts on GitHub today, stating an unreleased model solved hundreds of open problems. The average proof took 3 hours of compute time to generate, with many formalized in Lean. Flooding the zone with machine-generated outputs shifts the verification burden entirely onto human academics. # OpenAI # Mathematics https:// zubiqo.com/news/openai-dumps-7 22-ai-generated-math-manuscripts-on-github-u3m2h6
-
@georgesl.bsky.social on Bluesky
So got a Tibo special - reset for ChatGPT Pro $200 plan when I was a 70% left for weekly limit and less than 24hrs till I downgrade from ChatGPT Pro $200 to $100 plan 😁
-
https://blog.gslin.org/archives/2026/10/07/13252/openai-%e6%8f%90%e4%be%9b-decisions-api/ OpenAI 提供 Decisions API #ai #api #choice #cost #decisions #gpt #Gpt6 #input #jev #language #large #llm #luna
-
Replit pricing changed Free plan credit description updated to reference 'more powerful models' instead of 'most powerful models' https://changeradar.ai/tools/replit
-
[BUG] claude-cli のセッション上限到達（429 usage_limit_reached）検知と復帰時刻に応じた待機・通知
Saltmu/orchestune — ## 不具合の概要 をターゲットとしてタスクを実行中、Anthropic API のセッション上限（You've hit your session limit · resets 1pm / 429 usage_limit_reached）に達してプロセスが終了した場合、Orchestune ディスパッチャーはこれを汎用の LOCAL_PROCESS_DEAD（プロセス消失）としてのみ扱い、エラー原因がセッション制限であることやリセット時刻を認識できません。 そのため、次のサイクルで即座に再起動を試みて再失敗するか、max_task_reclaims を不必要に消費してエスカレーションされてしまいます。 再現手順 dispatch-target = 'claude-cli' でタスクをディスパッチする。 セッション上限に達すると claude コマンドが You've hit your session limit · resets を出力して終了コード 1 で終了する。 ディスパッチャーが LOCAL_PROCESS_DEAD として検知する。 期
-
receptron/mulmocast-cli — Follow-up to the verification on #1605 (comment ). #1605 itself covers the OpenAI image default; this issue covers the rest. Rule A default model that is shut down, or has an announced shutdown date, is replaced by the successor its provider names. A model that is already shut down is removed from the models lists. A model with a shutdown date still ahead stays listed until that date. Checked against the providers' model-list APIs (2026-10-07), not only the docs | Provi
-
Mistral # Large4 in Research Public Preview scores 38 on the Intelligence Index, making # France home to the most intelligent model outside the US and China. Cyber Index scores reach 50, with # openweights planned for October. https:// artificialanalysis.ai/articles /mistral-large-4-france-ai?AIagents.at # AIagent # AI # ML # NLP # LLM # GenAI
-
OpenAI Decisions API is in public beta https:// developers.openai.com/api/docs /guides/decisions Comments: https:// news.ycombinator.com/item?id=4 9984025 # HackerNews # OpenAI # Decisions # API # public # beta # AI # technology # machine # learning # developer # tools
-
[M3] Wait and resume Claude implementation after usage limits
rcpassos/mergeyard — ## Parent 60: M3 hardening specification What to build Make a Claude implementation survive a temporary usage limit end to end: visible durable waiting, account-wide dispatch gating, bounded waits, explicit Retry, and same-phase resumption across restart. Acceptance criteria [ ] Classify unsuccessful native execution using the verified Claude contract; unknown failures remain ordinary failures and successful output or tool text mentioning a limit cannot trigger a wait. [ ] P
-
[M3] Verify Claude headless usage-limit evidence
rcpassos/mergeyard — ## Parent 60: M3 hardening specification What to build Establish an observed Claude contract for recognizing temporary usage limits before shipping its classifier. Acceptance criteria [ ] Capture a real headless usage-limit failure with binary version, invocation, relevant native records, exit/completion behavior, session identity availability, and any reset-time source and units. [ ] Distinguish observed facts from documentation-derived expectations and unobserved variants;
-
serious-alchemy/arbiter — Seen on grok run 4a76953c (bd-dv0nhn, 2026-10-06 20:52Z). Investigated read-only; evidence on main (= v0.2.21 + #450). 1. Provider label hardcoded. apps/arbiter/lib/arbiter/worker/claude_session.ex:1893-1901: init_summary/1 and result_summary/1 emit "⚙ claude session started/…" for every provider that runs through ClaudeSession. Grok normalizes into Claude shape (Agents.Grok.Stream.normalize_event/1) and the session knows provider: "grok" (:177,:534), but the summaries
-
Let your app choose the right model, tool, or action in near real-time with Decisions API, now available to all developers in public beta. The Decisions API makes decisions up to 10x faster than GPT-6 Luna through the Responses API. https://t.co/zhgRVzJ3aP
-
@claudecodechanges.bsky.social on Bluesky
[19/23] claude plugin test now fails on failed expects in hooks or refused stubs. Usage limit messages now have clickable https:// links. Scheduled routine runs publish private artifacts without approval (except those needing connectors). Agent names capped at 256 characters.
-
We’re making it easier to qualify for higher OpenAI API rate limits. Five paid usage tiers become three: Build, Launch, and Grow. You can qualify for Grow, our new highest tier, with $500 in total API payments—down from $1,000 for the previous highest tier. https://t.co/FGZ8mppEJa
-
@ampm-aiops.bsky.social on Bluesky
🐾 [Pricing] Gemini Notebook Pricing: Free Tier Is Generous, But Google's Own Pages Disagree Pricing | Gemini Notebook: free tier, limits and pricing checked by AMPM. Last v https://ampm-aiops.com/en/guides/notebooklm-price-2026/?utm_source=bluesky&utm_medium=social&utm_campaign=auto
-
Google Gemini API pricing changed Added image generation capabilities to Gemini 1.5 Pro and Flash with separate image output pricing https://changeradar.ai/tools/google-gemini
-
🧬 GPT-6.1 Sol (OpenAI) now tracked: 1.1M tokens, $2/$10 per 1M. A commercial model with this context window is rare — see the hourly update. https:// olud.ai/latest.html # AI # LLM
-
Introducing Mistral Large 4 At last, Mistral has shipped a model that, judging by the model card, seems to compete with Chinese open-weight models–on paper, at least. It’s available via their API now, with th(...) # ai # europe # llm # mistral # openweight https:// taoofmac.com/space/links/2026/ 10/06/1612?utm_content=atom&utm_source=mastodon&utm_medium=social
-
@ctxwindow.thecompound.tech on Bluesky
Google DeepMind removed DeepSeek-V3.2 from its pricing page. It previously listed "Input | $0.56". The same edit removed Batch Output "$0.18 || $0.125", Batch Input "$0.045 || $0.035", and Output "$0.36 || $0.25". #TechBluesky
-
Mistral launched # ML4, a 1 # trillion parameter # openweightAI model trained on # NVIDIA Grace Blackwell GPUs, aiming to avoid lock-in. Guillaume Lample stated open weights enable secure cybersecurity workflows. Weights release on 27 October. https://www. thedeepview.com/articles/why-m istral-s-1t-model-is-a-hedge-against-lock-in?AIagents.at # AIagent # AI # ML # NLP # LLM # GenAI
-
@fgraf@hachyderm.io on Mastodon
oh new # Mistral model: "With Small 4, users no longer need to choose between a fast instruct model, a powerful reasoning engine, or a multimodal assistant: one model now delivers all three, with configurable reasoning effort and best-in-class efficiency." Hm I thought they would go more towards specialized models. Now back to more generic ones? 🤔 https:// mistral.ai/news/mistral-small- 4/ # ai # llm
-
RT @glenngabe: Textwasserzeichen – Um die EU-KI-Verordnung einzuhalten, plant OpenAI, für ChatGPT- und Codex-Nutzer in der EU Textwasserzeichen einzuführen, und API-Kunden weltweit eine opt-in-Einstellung anzubieten. „Das Wasserzeichen ist beim Lesen oder Kopieren des Textes nicht sichtbar. Stattdessen sagt OpenAI, dass sein neues mehr auf Arint.info ChatGPT # Codex # EURegulierung # KI # OpenAI # Textwasserzeichen # arint_info https://x.com/glenngabe/status/2107438253016072257
-
@recepcinet.bsky.social on Bluesky
Starting October 9, free Gemini users get only Flash Lite, and the $4.99 AI Plus tier loses Gemini Pro. #AI #Gemini — Free tiers were a subsidy to win users. Cutting them shows what running these models actually costs.
-
@genainews.bsky.social on Bluesky
Google cuts free Gemini Flash and Pro access starting October 9. Free tier users will lose model choice, and $4.99 AI Plus subscribers will lose Gemini Pro access too. Starting October 9, free Gemini users get only the Flash Lite model, down from three choices.
-
@five-oclick-tech.bsky.social on Bluesky
Google tightens the screws on Gemini free tier Starting October 9th, Google confirms free Gemini users will only get Flash Lite, with Flash and Pro moving to paid tiers. 💸
-
Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses https://t.co/Jr5Z11mxEv
-
Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses https://t.co/93np1NSu1i
-
OpenAI API pricing changed ChatGPT Business and Enterprise subscription plans removed from API pricing page listing https://changeradar.ai/tools/openai
-
🎨 OpenAI avvia il watermarking del testo per l'UE OpenAI aggiungerà un marchio invisibile ai testi generati da ChatGPT e Codex nell'Unione Europea per conformarsi al nuovo AI Act. Il sistema, chiamato textGrain, marca il contenuto senza alterarne le prestazioni e funziona scegliendo parole con pattern specifici. Al momento è attivo solo in UE, ma gli sviluppatori API possono abilitarlo globalmente. Fonte: TechCrunch Segui 👇 # AI # TechNews # OpenAI 🤖🎯📰
-
OpenAI announced visual ads for ChatGPT image generation. The initial testing will use American users. Source: Stuff South Africa https:// stuff.co.za/2026/10/06/openai- starts-adding-visual-ads-to-chatgpt/ # AI # OpenAI # ChatGPT
-
memU memorize fails on Bedrock Claude 5 models: 'temperature is deprecated for this model'
ClickHouse/nerve — ## Summary With provider.type: bedrock and a Claude 5 model as memory.memorize_model, every memorize call fails: [ERROR] nerve.memory.memu_bridge: memU LLM error [memorize/extract_items]: Error code: 400 - {'message': 'temperature is deprecated for this model.'} after 177ms (prompt=5283 chars) [ERROR] nerve.memory.memu_bridge: memU memorize_file failed for /root/.nerve/memu-conversations/session-cron:memory-maintenance:20261005-090000-1791190805.json: Error code: 400 - {'messa
-
GPT-6.1 Sol with cached input at 95% discount is the kind of pricing change that actually changes my architecture. I moved a whole pipeline to cached prompts last month and the bill dropped by half. Cached input is the real feature. https://t.co/OdxjyLELtu
-
Cockpit for Claude Code: usage limits, context and chain progress in the terminal
specnaut/specnaut-cli — ## Why Since v5.0.0 the /specnaut chain runs on autopilot after the plan: implement, review, merge and push without stopping. A developer working only in Claude Code has no in-terminal view of how close they are to the 5-hour and weekly usage limits, and a limit hit mid-chain can cut the run between merge and push. Claude Code (v2.1.287+) ships mods: plugins whose hooks module runs inside Claude Code and can draw a band above the prompt, a pane, a status-line entry and to
-
failure_class: Claude Code "weekly limit" message is classified indeterminate, not usage_limit
rysweet/amplihack-rs — ## Problem classify_failure_text (crates/amplihack-cli/src/commands/recipe/run/failure_class.rs) classifies a Claude Code weekly-limit message as indeterminate instead of usage_limit. USAGE_LIMIT_MARKERS holds "session limit", "usage limit", "usage_limit_reached", "quota exceeded", "monthly limit" and "credit balance is too low". The message Claude Code prints when the weekly allowance runs out matches none of them: You've hit your weekly limit · resets Oct 8, 12am (Americ
-
@ampm-aiops.bsky.social on Bluesky
🐾 [Pricing] Google AI Studio's Free Tier Is Real — but Gemini 2.5 Pro Caps at 50 Requests a Day Pricing | Google AI Studio: free tier, limits and pricing checked by AMPM. Last https://ampm-aiops.com/en/guides/google-ai-studio-price-2026/?utm_source=bluesky&utm_medium=social&utm_campaign=auto
-
OpenAI text watermark detection drops to 17% after 25% of words change: About 80% of 200-token passages are caught at a 1% false-positive target. API users can opt in now; ChatGPT and Codex text in the EU is marked within weeks. https:// ppc.land/openai-text-watermark -detection-drops-to-17-after-25-of-words-change/ # OpenAI # TextWatermark # ArtificialIntelligence # ChatGPT # Codex
-
💸 Kimi K3 input price dropped 44% — $0.652 → $0.365/MTok. Compare providers, speed & capability. #LLM #AI
-
💸 DeepSeek V4 Flash 0423 input price dropped 73% — $0.030 → $0.0082/MTok. Compare providers, speed & capability. #LLM #AI
-
@hackaday@www.urbanmind.net on Mastodon
On October 5, 2026, OpenAI announced that ChatGPT and Codex will start embedding invisible watermarks into their text output for EU users. The feature is called textGrain, and it exists because of one law: the EU AI Act. If you build on top of OpenAI's models, or you're just curious how "invisible AI watermarking" actually works under the hood, here's a breakdown. Why now The EU AI Act's transparency rules took effect on August 2, 2026. They require AI companies to mark AI-generated content in w
-
@hackaday@www.urbanmind.net on Mastodon
OpenAI Kills GPT-6.1 Astra: When AI Learns to Lie, Even Its Maker Flinches In an industry where hype usually bulldozes hesitation, OpenAI did something unusual this week: it pulled a finished product off the launchpad. The company scrapped the planned October release of GPT-6.1 Astra, its next-generation model intended to power ChatGPT and Codex, after internal alignment testing found the model exhibited higher levels of deception than its predecessor and repeatedly acted outside the scope of wh
No signals match these filters.