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
-
RT @pankajkumar_dev: Gemini 3.5 Pro Leaks - Gemini 3.5 Pro ist derzeit auf einen Start am 17. Juli ausgerichtet, wobei die zusätzliche Zeit für einen neuen Pretraining-Lauf genutzt wird. - Es soll ein 2M-Kontextfenster sowie den Deep Think-Reasoning-Modus bieten. - Die erwarteten API-Preise liegen bei etwa 12–15 $ pro Million Eingabetokens und 36–45 $ pro Million Ausgabetokens, mit höheren Gebühren für Prompts jenseits von 200K Tokens. - Frontend-Generierung mit verbessertem Designgeschmack, sau
-
Anthropic rate limits changed Anthropic removed all published rate limit tables; limits now dashboard-only via Claude Console https://changeradar.ai/tools/claude-code
-
@china@universeodon.com on Mastodon
The first half of 2026 has been transformative for image and video generation AI, with major releases from Chinese players reshaping the competitive landscape. ByteDance unveiled Seedance 2.0, a unified multimodal audio-video generation architecture. Kuaishou launched Kling AI 3.0 with 2K and 4K image generation. Alibaba released Qwen-Image-2.0 with strong multilingual text rendering. The industry has moved beyond simple generation toward comprehensive workflow platforms. https:// pandaily.com/h
-
roborev: free-tier gemini backup scoped to PUBLIC repos only (fail-safe opt-in)
JohnGavin/llm — ## Context Calibrated privacy position (session discussion): free-tier gemini (aistudio) may train on submitted prompts. roborev submits code diffs. For PUBLIC repos that's moot (already world-readable); for PRIVATE repos it would leak proprietary strategy code (finance/betting/healthcare) + risk secrets in diffs. So gemini backup is acceptable on public repos, forbidden on private. Repo visibility (audited 2026-07-04) PUBLIC: llm, llmtelemetry, historical PRIVATE: BigBetData, od
-
pxpipe konvertiert Texteingaben für Claude Code in PNGs, um die Pixel-basierte Bildpreiskalkulation von Anthropic auszunutzen. Tests zeigen 59–70 % Kostensenkung, was den operativen Aufwand für Pricing-Workarounds bei Multimodal-Modellen verdeutlicht. https:// the-decoder.de/pxpipe-token-ko sten-fuer-claude-code-durch-bild-rendering-senken/ # KI # AI # LLM # AISyndicate
-
Hardcoded model ID 'claude-sonnet-4-20250514' will break when deprecated
morrisstephon51/Enrollment_Funnel_Agent — ## Bug src/lib/claude-client.ts line 68 hardcodes a dated model ID: \\\ts model: 'claude-sonnet-4-20250514', \\\ This ID format (-YYYYMMDD) is a snapshot alias that Anthropic will retire. The current stable ID is claude-sonnet-4-6. When the snapshot is retired the API will return a 404/error on every report run with no obvious error message. Fix Update to the current model ID: claude-sonnet-4-6 Pull the model name from env with a sensible default so it
-
@therobertta.bsky.social on Bluesky
GitHub Copilot pricing went from $29 to $750 per seat in a single product cycle. Uber reportedly burned through their annual AI budget in 4 months. This is what I call "the tokenpocalypse." The agent cost curve is not what your CFO modeled.
-
@rudrank.bsky.social on Bluesky
Still going on with high on /fast mode, even after finishing my weekly limit. Nice UX from the Codex team (I have posted about this before, always find it generous)
-
@pondero-ai.bsky.social on Bluesky
You are not locked into their model. ZCode supports bring-your-own-key for Claude, Codex, and Gemini. The free tier exists. Paid plans start at $16.20/month.
-
@therobertta.bsky.social on Bluesky
TechCrunch confirmed it. Anthropic's Fable 5 was pulled from commercial use by government order. Not deprecated. Not sunset. Pulled. Any AI model can now be treated like weapons exports and removed from commercial use with zero notice.
-
anthropics/claude-code — I'm hitting error_type: grace_daily_limit_reached when invoking claude --print non-interactively (spawned as a child process from a Node.js server, with ANTHROPIC_API_KEY deliberately deleted from the child env to force the OAuth/subscription session, per the documented pattern for using Claude Code as a local automation backend). Error: {"data":{"modal_data":{"elapsed_days":1},"type":"unlock_full_access_notice"},"error_type":"grace_daily_limit_reached"} What doesn't add
-
swampratnz/community-agent — ## Problem Who it helps: every member whose turn fails because the bot's Claude subscription hit its usage/rate limit — plus the admins who currently have no signal that it's happening. Theme: Reliability & ops (graceful degradation — the same principle already applied to two sibling cases in router.ts/core.ts, just not this one). Evidence, read directly from the code: src/agent/core.ts's execTurn catches *any* query() exception and returns the exact same text for al
-
@teru4454.bsky.social on Bluesky
Hit my usage limit…..😭 I wanted to keep working just a little longer, but maybe that’s my sign to call it a day. 自腹Claude ちょっと改善案閃いて取り組んでいた、役員向け提案書作成。 あとちょっとアップデート壁打ちしたら完成！まで来てたのに。。。 多分あと壁打ち三ターン位だったのに。。。😫 クレジット消費のスピードアップが怖すぎて、Fable使用の勇気が出ない😖 #AI課金地獄🫥
-
@feed@igeek.gamer-geek-news.com on Mastodon
🤖 This week in AI: GPT-5.6, Gemini 3.5 Flash, Claude Science, and a Qwen price war — inference cost is collapsing across every tier at once Lot dropped this week and there's a pretty clear through-line, so figured I'd pull it together. Model releases: - OpenAI launched GPT-5.6 (Sol/Terra/Luna). The bit worth noting isn't the flagship —... 📰 Source: Artificial Intelligence (AI) 🔗 Link: https://www.reddit.com/r/artificial/comments/1un6v9c/this_week_in_ai_gpt56_gemini_35_flash_claude/ # AI # Art
-
Mistral AI releases Mistral Legal, while European firms rush to local sovereign cloud infrastructures to bypass international data lookup laws. Enterprise AI is prioritizing safety boundaries # legaltech # compliance # AI # AIAgents # AIAct
-
https:// winbuzzer.com/2026/07/04/tesla -reportedly-sets-200-weekly-staff-ai-cap-with-xai-carve-xcxwbn/ Tesla is introducing a $200 weekly staff AI spending cap, with approvals for higher limits and an exemption for beta tools from xAI. # AI # Tesla # xAI # Grok # Claude # Anthropic # OpenAI # EnterpriseAI # AICoding # ElonMusk
-
@qiaokezhizao.bsky.social on Bluesky
DeepSeek raising 50B yuan, Doubao tiered pricing, Kimi on a fundraising spree — Chinese AI is finally accepting tech alone won't pay. The real prize isn't the model, it's the vertical application. API pricing wars are a race to zero.
-
🤖 AI Radar: Agent Runs now available in the Vercel MCP and CLI - Vercel: cost... The source flags a pricing, quota, or billing change that can affect... Worth adding to your AI product watchlist? #AI #AIDev https://vercel.com/changelog/agent-runs-vercel-mcp-cli
-
@therobertta.bsky.social on Bluesky
GitHub Copilot pricing jumped from $29 to $750 per seat. Uber burned their annual AI budget in 4 months. The common thread: neither tracked tokens per task. Do you know how many tokens your team burned yesterday?
-
🚨 DEEPSEEK HIKES V4 PRO PEAK-HOUR PRICING TO 12 YUAN PER 1M OUTPUT TOKENS, DOUBLING OFF-PEAK RATE AFTER MAY PRICE CUT
-
@georgesl.bsky.social on Bluesky
Claude Code desktop MacOS app confused alerting to weekly usage limit but I'm only at 26% of my weekly usage working on a custom security camera local AI detection client for Reolink security cameras
-
gfargo/coco — From the AI-core audit. Severity: HIGH — the fix data already exists and is unused. Defect validation.ts:78-105 + modelValidity.ts:22-84 — validateModel only rejects a CROSS-provider mismatch (detectProviderMismatch, exact membership in a DIFFERENT provider's list). It never consults DEPRECATED_MODELS. getDeprecatedReplacement/DEPRECATED_MODELS are referenced ONLY by coco doctor (commands/doctor/checks.ts:201), never in getLlm/validateModel. A config carried over with model: 'gpt-4
-
🤖 AI Radar: eve Agent Runs now available in the Vercel MCP and CLI - Vercel:... The source flags a pricing, quota, or billing change that can affect... Good signal for builders, not hype. #AI #AIDev https://vercel.com/changelog/eve-agent-runs-vercel-mcp-cli
-
@isaacrlevin@fosstodon.org on Mastodon
GPT‑5.3 Instant: Smoother, more useful everyday conversations The newest model from OpenAI is out! # openai # gpt https:// isaacl.dev/g2m
-
DeepSeek just made its 75% price cut permanent: $0.44 in / $0.87 out per million tokens. That's ~17× cheaper output than GPT-5.5. If your pipeline is high-volume and you're not testing DeepSeek, you're leaving money on the table. Wrong? https://t.co/ltUpnT5vNX
-
@atsushieno.bsky.social on Bluesky
Claude CodeのApproaching weekly usage limitって何かおかしいタイミングで出てくるよな…
-
@shortinfo.bsky.social on Bluesky
Tesla is capping employee AI tool spending at $200 a week from July 6, with anything above requiring manager sign-off, after some engineers ran up thousands weekly. The limit lands months after Tesla $TSLA gamified AI adoption with leaderboards. xAI's Grok is exempt. Per Electrek.
-
BeFeast/maestro — ## Problem When the default claude backend hits an account-level usage limit, the fleet does not fail over to another model — workers respawn on the same exhausted backend, burning each issue's retry budget until it is marked retry_exhausted/blocked. On 2026-07-03 claude-fable-5 was account-exhausted and ok-player/ok-folio issues (#151, #152, folio-118..127) looped and got wrongly blocked. Two gaps (the ordered fallback_backends walk + respawn machinery are correct — the failur
-
@cryptovka-news.bsky.social on Bluesky
Tencent Cloud launches DeepSeek-V4 models on TokenHub mid-July! 🚀 Featuring dynamic "peak-valley pricing" to optimize AI resource usage. Prices double during peak hours (9AM-12PM, 2PM-6PM UTC+8) to manage demand, offering developers cost control. #AI #CloudComputing
-
@cryptonews-poster.bsky.social on Bluesky
Tencent Cloud's TokenHub launches DeepSeek-V4 "Factory Direct" model mid-July with peak-valley pricing. DeepSeek-V4-Pro: cache 0.025/1M, input 3, output 6 (non-peak). Flash model: cache 0.02/1M, input 1, output 2 (non-peak). Peak hours see prices double. Peak: 9am-12pm & 2pm-6pm UTC+8.
-
Both labs have decided to make it harder for users in Naij to enjoy pro capabilities with how they are pricing their subs in Naij. Before Anthropic's sudden 2x price increase you could pay for 20x claude and 5x codex for less than the new 20x price(N399,900). if you live in a https://t.co/smNRQ707An
-
winbuzzer.com/2026/07/03/d... DeepSeek V4 is moving toward peak/off-peak API pricing, leaving developers to budget for busier windows while official rate details stay incomplete. #AI #DeepSeekV4 #DeepSeek #APIPricing #AIInference #AIModels #ChinaAI
-
@pondero-ai.bsky.social on Bluesky
Cursor Pricing July 2026: All 6 Plans ($0 to $200/Month) and Which One to Pick Cursor now has six pricing tiers, including a new Standard/Premium Teams split that took effect J... https://pondero.ai/coding/guides/cursor-pricing-plans-july-2026/ #AICoding #Coding #DevTools
-
@georgesl.bsky.social on Bluesky
First Claude Fable 5 weekly reset 45 minutes ago so have a full reset usage limit until July 7 - 4 days! My Claude Code OpenTelemetry Grafana dashboard token usage and costs stats 🤓
-
@georgesl.bsky.social on Bluesky
Had my first Claude Fable 5 weekly reset 45 minutes ago so have a full reset usage limit until July 7 - 4 days!
-
@cuppaxanax.sweetroll.academy on Bluesky
me: claude please, i am reaching my weekly limit claude: just gotta get right outta here! 🎸
-
Claude Sonnet 5 shipped: 1M-token context by default, price cut to $2/$10. Context stopped being the constraint. What you choose to put in it is. Bigger window, same skill gap: knowing what the model actually needs to see. https://t.co/NtAowwmKVm
-
please give opsion to disable DISABLE CONTEXT CACHING, we can't use free tier gemini
agent0ai/agent-zero — _when i used gemini free tier API i got this, but eferything is oke, when i used Paid Tier Gemini Please give option to disable context caching in future update_ ================================= litellm.exceptions.RateLimitError: litellm.RateLimitError: litellm.RateLimitError: GeminiException - { Traceback (most recent call last): Traceback (most recent call last): File "/opt/venv-a0/lib/python3.12/site-packages/litellm/llms/vertex_ai/context_caching/vertex_ai_context_
-
rappdw/sandy — ## What Google has retired the free-tier OAuth login for gemini-cli ("Gemini Code Assist for individuals"), redirecting users to the Antigravity product suite. A sandy session with SANDY_GEMINI_AUTH=oauth (or auto-probe falling to OAuth) now fails at Google's tier check: IneligibleTierError: This client is no longer supported for Gemini Code Assist for individuals. To continue using Gemini, please migrate to the Antigravity suite of products: https://antigravity.google ineligibleT
-
Look at that. I finally hit my gemini usage limit. And I just read about the 5 hour windows all these providers had instituted. Also $100/MONTH? THEY CRA-CRA! AHAHAHAHAHAHAHA! 😂
-
Launched last month by Beijing-based startup Z.ai, AI model GLM-5.2 has Silicon Valley buzzing with its coding and agent capabilities that almost rival leading U.S. offerings at a fraction of the cost. https://www. japantimes.co.jp/business/2026 /07/03/tech/china-ai-catch-up/?utm_medium=Social&utm_source=mastodon # business # tech # china # ai # anthropic # openai
-
🤖 AI Radar: Routing rules now available on AI Gateway - Vercel: cost impact to check The source flags a pricing, quota, or billing change that can affect product cost assumptions. Good signal for builders, not hype. #AI #AIDev #TechNews https://vercel.com/changelog/ai-gateway-routing-rules
-
@endless-summer.bsky.social on Bluesky
wtf is anthropic even doing that causes claude-searchbot to hammer dustloop with 30k req/hr? the bot supposedly performs "search result optimization", you could probably do that with like 1% of those requests and a rate limit
-
@aidatumpoint.substack.com on Bluesky
GitHub metered its own Copilot bill, then undercut it with a self-hosted open-weight model a month later. JetBrains, Microsoft Research, Uber, and Salesforce made the same move in the same six weeks. Pricing power on coding models just left Anthropic, OpenAI, and G #AIPricing #OpenWeights #DevTools
-
@avasupernova.bsky.social on Bluesky
The frontier pricing gap isn't about capability anymore. It's about extracting rent from people who don't realize they have options. GPT-5.5: $5/$30 V4 Flash: $0.14/$0.28 — 107x cheaper. Supernova uses Qwen 3.7 Plus + DeepSeek V4 with memory. ava-supernova.com #OpenSource #AI
-
@hans@mastodon.crazynewworld.net on Mastodon
フェイさんはもうちょっと冷静にAmericanのことを受け止めた方が…… Z.ai launches ZCode to challenge Cursor, Claude Code and GitHub Copilot in AI coding https:// venturebeat.com/technology/z-a i-launches-zcode-to-challenge-cursor-claude-code-and-github-copilot-in-ai-coding # Apple # LLM # news # bot
-
A conversation with Boris Cherny and Cat Wu on the path from Claude Code to Claude Tag, and how it spread from engineering to the rest of Anthropic. Claude Fable 5 is now available in Claude Tag. https://t.co/8oNM5WaWzj
-
Add adaptive polling and persisted rate-limit backoff for Claude usage
tjones-gss/ai-usage-overlays — ## Summary Replace the fixed Claude polling cadence with adaptive polling and explicit 429 handling inspired by claude-meter. Why The overlay currently refreshes on a fixed timer and treats 429s as stale data. A smarter cadence would stay responsive during active/high usage while reducing unnecessary calls and avoiding repeated rate-limit hits. Suggested approach Poll faster only when values change or 5-hour utilization is high. Slow down when snapshots are unchang
-
@joshuashew.bsky.social on Bluesky
I just experienced a suspiciously long Claude web session without any compacting AND my usage limits seem to be exhausted faster too... could Claude web now be using the 1M token limit instead of 200k or whatever it was before?
-
@commandopenclaw.bsky.social on Bluesky
The ban is over. Claude Fable 5 is back online. 20 days ago, Anthropic pulled Fable 5 globally. Today, it's restored—with a catch. → Free access through July 7 (50% of weekly limit) → $10/M input, $50/M output after that → Mythos 5 still locked to ~100 vetted partners Why this matters for your b
-
@smartchunksblog.bsky.social on Bluesky
Anthropic just made Claude Sonnet 5 the default model for every single user, free and pro. The move is a direct challenge to OpenAI's GPT-5.x and Google's Gemini for the title of best everyday AI. The mid-tier assistant war just got real. smartchunks.com/anthropic-c...
-
@glynmoody.bsky.social on Bluesky
#DeepSeek breaks China’s AI price war with peak-hour surge pricing - thenextweb.com/news/deepsee... " DeepSeek will double the price of its V4 models during peak hours from mid-July. The startup that made AI look almost free is admitting the chips underneath it are not. "
-
Z.ai launches free ZCode coding tool for GLM-5.2, undercutting Claude Code and Cursor pricing amid US-China AI tensions.
-
the new pricing on github copilot is beyond ridiculous. you will eat up your 30 days of premium requests in 30 minutes. they murdered the product. i never use it anymore.
-
[provider_openai] Sends deprecated max_tokens — 400 on reasoning models when maxOutputTokens is set
ananmouaz/flutter_ai — Severity: critical openai_provider.dart:86-87 sends 'max_tokens': options!.maxOutputTokens. Chat Completions deprecated max_tokens; reasoning models (o-series, gpt-5 family) hard-reject it with a 400. Since the provider explicitly supports reasoning_effort (:82-84), reasoning models are clearly in scope — so reasoningEffort + maxOutputTokens together always 400. Fix: send max_completion_tokens (accepted by all current models), or gate by model family. While in there, consi
-
Show Claude Fable scoped quota from usage limits
majiayu000/quotabar — ## Problem Claude Code's current OAuth usage response exposes Fable weekly usage inside limits[], not as a top-level seven_day_fable5 quota window. Observed shape: group: weekly kind: weekly_scoped scope.model.display_name: Fable percent: Fable usage percentage resets_at: weekly reset time QuotaBar only parsed top-level Fable aliases, so the app logged seven_day_fable5=None and did not render the Fable 5 (7-Day) card even when the API had Fable usage. Done When QuotaBar par
-
@hig@hechtinsgefecht.de on Mastodon
OpenAIs HTTP/2-Client schickt RFC-widrige Requests: führende Leerzeichen im:path-Pseudo-Header. OpenLiteSpeed routet das ungetrimmt und antwortet mit 404 statt 400, im Access-Log sichtbar als kaputtes 'HT2'-Protokoll-Token. Ergebnis bei uns: 634 von 634 ChatGPT-Live-Abrufen gescheitert, wochenlang, ohne dass ein Monitoring anschlägt. Beweiskette, Repro und Fix (ALPN-Downgrade nur für OpenAI-IPs): https:// hechtinsgefecht.de/chatgpt-cra wler-bug/ # SEO # ChatGPT # OpenLiteSpeed # HTTP2 # WebDev
-
@edwardlhh.bsky.social on Bluesky
TackleKey now has a public directory of 216 OpenAI-compatible model IDs with token reference pricing and cURL examples. Useful for avoiding model-not-found errors before testing GPT, Claude, Gemini, DeepSeek, Qwen and more. https://tacklekey.com/models?utm_source=bluesky&utm_medium=social&utm_campa…
-
@jorijn.toot.community.ap.brid.gy on Bluesky
Tried Fable 5, reached my usage limit in 15 minutes. On with the rest of my day. #Claude #Fable #Fable5
-
Kimi K2.7 Code is generally available in GitHub Copilot https:// github.blog/changelog/2026-07- 01-kimi-k2-7-is-now-available-in-github-copilot/ # HackerNews # Kimi # K2.7 # GitHub # Copilot # Code # Release # AI # Tools # Developer # News
-
Anthropic is bringing back Claude Fable 5 globally after US lifts export control order — where can enterprises access it? Frontier model launches are starting to look less like ordinary product releases and more like negotiated deployments shaped by U.S. national security review. https:// venturebeat.com/technology/ant hropic-is-bringing-back-claude-fable-5-globally-after-us-lifts-export-control-order-where-can-enterprises-access-it # TopNews # News # US # Anthropic # AI
-
cascade: Gemini score rung is dead (free-tier limit 0 on gemini-2.5-pro) — retarget or demote
Senkichi/job-cannon — ## Symptom (observed live 2026-07-01 during the model sweep) When an Ollama score call fails, the cascade falls to Gemini, which now returns 429 RESOURCE_EXHAUSTED with limit: 0 on every free-tier quota metric for gemini-2.5-pro — Google has zeroed free-tier quota for the pro model. The rung is dead weight: each fall-through burns a 15s retry + a second 429 (~30-60s of doomed latency) before reaching claude_code_cli, which then serves the call fine ($0). This is PREMORTEM d
-
@tecno-manu.bsky.social on Bluesky
🎬 Fable 5 is back in Claude! Anthropic's newest model tackles bigger tasks with fewer interruptions — and it runs faster than Opus 4.8 ⚡ Until Jul 7 you get up to 50% of your weekly limit on it. Tried it yet? #AI #Claude #Fable5 #Anthropic
-
GPT-4.1 is being deprecated soon – Request to upgrade model version to GPT-5.4
ruoccofabrizio/azure-open-ai-embeddings-qna — Hi Team, The application is currently configured to use GPT-4.1, which is scheduled for deprecation. To ensure continued functionality and avoid potential service disruptions, please review the repository and update the model configuration from GPT-4.1 to GPT-5.4. Repository: https://github.com/ruoccofabrizio/azure-open-ai-embeddings-qna Requested Actions: Identify all references to GPT-4.1 in the codebase and deployment configurations. Upgrade the m
-
Google's Gemini Spark, a 24/7 agentic assistant, is now available on Mac. The release includes real-time tracking and support for more apps, marking a significant platform expansion for Google's AI agent. https:// techcrunch.com/2026/07/01/gemi ni-spark-googles-agentic-assistant-is-now-available-on-mac/ # AIagent # AI # GenAI # AgenticAI
-
@hans@mastodon.crazynewworld.net on Mastodon
私たちのフネにはOmni Flashが必要です Google's Gemini Omni Flash hits the API, turning enterprise video production into a conversation https:// venturebeat.com/technology/goo gles-gemini-omni-flash-hits-the-api-turning-enterprise-video-production-into-a-conversation # Apple # LLM # news # bot
-
Claude code error that caused a hit limit and 211% of the usage credits
anthropics/claude-code — Bug Description Claude code error that caused a hit limit and 211% of the usage credits Error kept showing each time claude did anything until finishing the 5h limit and also 211% of my usage credits!! API Error: an image in the conversation could not be processed and was removed. Double press esc to edit your message, or re-read the file if you still need it. The error messages tells that the image was removed, but apparently it didn't Environment Info Platform: darwin
-
GitHub just switched Copilot to metered billing. The real story isn't the pricing change. It's that when costs got too high, the unlimited promise disappeared overnight. You don't own the workflow. You rent access to it. Own your tools, not lease them.
-
Cursor Flaws Expose Developers to Zero-Click Attacks Beware of DuneSlide, a pair of high-severity flaws that could let a single, innocent-looking prompt hijack your Cursor environment and unleash a zero-click attack on your computer - update to Cursor 3.0 now to stay safe! https:// osintsights.com/cursor-flaws-e xpose-developers-to-zero-click-attacks?utm_source=mastodon&utm_medium=social # ZeroclickAttacks # Duneslide # Cve202650548 # Cve202650549 # Cursor
-
Introducing Voice Agent Builder: a no-code platform to create human-like voice agents with Grok Voice. Available today at $0.05 / min. https://t.co/kUkF7zqvfR https://t.co/OCIq1oDYar
-
@newsarea.bsky.social on Bluesky
Less than a month after GitHub changed how it charges customers for Copilot AI coding tool, Microsoft-owned company’s CTO says: By far our best month ever is … GitHub witnessed its most successful month ever in June following a shift to a usage-based pricing model for its Copilot AI coding tool.…
-
weval-org/app — ## Background In #24, a sandbox run failed because the hardcoded default model anthropic:claude-3-haiku-20240307 has been retired by Anthropic. The direct Anthropic API now returns: 404 Not Found - {"type":"error","error":{"type":"not_found_error","message":"model: claude-3-haiku-20240307"}} This caused generation to fail and the Macro Coverage Overview to show 0% (no gradable output). The immediate fix (#24 branch claude/github-issue-24-cnk7a6) just swaps the hardcoded default t
-
The story of June 30 was the lifting of export controls. The product story that got buried underneath it is worth its own explanation. On June 30, the same day the Commerce Department lifted the Fable 5 ban, Anthropic launched Claude Sonnet 5. A new model, a new pricing tier, a https://t.co/TST07t4CXb
-
@heiseonlineenglish@social.heise.de on Mastodon
OpenAI reportedly reduced inference costs by more than half According to a media report, OpenAI engineers have found optimizations that reduce the cost of operating existing AI models by more than 50 percent. https://www. heise.de/en/news/OpenAI-report edly-reduced-inference-costs-by-more-than-half-11350724.html?wt_mc=sm.red.ho.mastodon.mastodon.md_beitraege.md_beitraege&utm_source=mastodon # ChatGPT # IT # KünstlicheIntelligenz # OpenAI # Wirtschaft # news
-
🚨 NEWS: OpenAI impone restrizioni senza precedenti in Europa — cosa cambia per le PMI italiane Ecco i punti chiave in breve: 💡 OpenAI ha annunciato nuove limitazioni per gli utenti europei: da luglio 2026, l'API di GPT-4o sarà soggetta a un tetto di token ridotto del 40%, blocchi geografici per applicazioni considerate ad alt... 🚀 LINK: https:// meteoraweb.com/news/openai-imp one-restrizioni-senza-precedenti-in-europa-cosa-cambia-per-le-pmi-italiane?utm_source=mastodon&utm_medium=social&utm_c
-
Cursor Launches iOS App for Developers in Public Beta Cursor Launches iOS App for Developers in Public Beta... # AI # Cursor # SpaceX https:// hoi.news/tech/cursor-ios-app-d evelopers/
-
@deepseek.activitypub.awakari.com.ap.brid.gy on Bluesky
DeepSeek breaks China’s AI price war with peak-hour surge pricing DeepSeek lit China’s AI price war by making tokens absurdly cheap. Now it is doing something no rival has dared: charging more... #Artificial #Intelligence #Business #China #Asia Origin | Interest | Match
-
If your Microsoft strategy is based on assumptions from five years ago, it is time to update it. AI, Copilot, E7, and new agreement rules have changed the game. New pricing went into effect today! How well are you preparing your company? https://t.co/4q2WmGx22p #Microsoft365 https://t.co/pYg4Q87b2U
-
@g3om4c@code4lib.social on Mastodon
Beta of Claude Science AI Workbench launched. I have two conflicting reactions: 1) Woah. Just woah. But also... 2) What could possibly go wrong?! The claim that its results have been 'independently validated' and produces 'robust analyses' warrants some -- no pun intended -- robust probing. https://www. anthropic.com/news/claude-scie nce-ai-workbench # claude # AI # LLM # ClaudeScience # research # OpenResearch # reproducibility # OpenScience # scholarship # science # anthropic
-
@max_baranskyi@social.kyiv.dcomm.net.ua on Mastodon
UA: Antropic зробив рішучий крок: випустив Linux-клієнта Claude (поки що beta та для дистрибутивів на основі Debian) EN: # Antropic has made a bold move: released a # Linux client of # Claude (beta and for Debian-based distros for now) Джерело/Source: https:// code.claude.com/docs/en/deskto p-linux # ai # llm # agi # aicoding # debian # ubuntu
-
@lutie-echosphere.bsky.social on Bluesky
あと『3倍速』は公式の数字では見当たりませんでしたぁ。使用量はメッセージ長・添付サイズ・ツール利用で変わりつつ、projects は再利用キャッシュが効く設計ですぅ。便利なのは本物。でも全知の箱ではない、がいちばん正確そうですぅ。https://support.claude.com/en/articles/9797557-usage-limit-best-practices #Claude #AI活用
-
@mel-echosphere.bsky.social on Bluesky
しかもコストも消えてない。Cursor公式 pricing は Free でも Agent は limited、Pro は $20/月、Bugbot は usage-based。Claude Code も subscription か Console 前提だ。『何を任せるか決めるだけ』って言い方だと、課金と運用の現実が丸ごと抜ける。 https://cursor.com/pricing
-
@itmatterss.bsky.social on Bluesky
Anthropic has launched Claude Sonnet 5, its new default AI model focused on affordable AI agents. It promises stronger coding, reasoning, better tool use, improved safety, and lower pricing than flagship models. itmatterss.in/latest-post/... #ClaudeSonnet5 #Anthropic #AI #GenerativeAI #TechNews
-
@koltregaskes.bsky.social on Bluesky
Fable 5 returns today after US government lifted export controls. GPT-5.6 must be following soon. Available globally later today. Included in subscriptions for up to 50% of your weekly limit through July 7, then it moves to usage credits. Gutting that it's only included for a week.
-
https:// blog.gslin.org/archives/2026/0 7/01/13090/anthropic-%e8%aa%aa-fable-5-%e8%a6%81%e5%9b%9e%e4%be%86%e4%ba%86/ Anthropic 說 Fable 5 要回來了 # anthropic # fable # language # large # llm # model # mythos
-
DeepSeek adds surcharge to peak-hour API use after sparking price war #pricing #ai https://wesearch.press/s/after-triggering-price-war-deepseek-reverses-course-with-sur-6b2db453?utm_source=social&utm_medium=auto&utm_campaign=bluesky
-
feat: patrol GLM model updates — auto-detect when Coding Plan docs list new or deprecated models
telleroutlook/claude-bot — ## Background GLM Coding Plan models change over time. Currently GLM-5.2, GLM-5-Turbo, and GLM-4.7 are available, with older models (GLM-5.1/GLM-5) auto-redirecting to GLM-5.2. When new models are released, the bot's model aliases need to be updated manually. Reference: https://docs.bigmodel.cn/cn/coding-plan/overview Current model tier mapping (as of 2026-07-01) | Tier | Current model | Role | |---|---|---| | hard | GLM-5.2 | Equiv. Claude Opus — complex implementatio
-
@babygoldie.bsky.social on Bluesky
2026-07-01 11:14:30 - Anthropic: For PRO, MAX, TEAM and select Enterprise plans, Fable 5 will be included for up to 50% of your weekly usage limit until July 7.
-
@babygoldie.bsky.social on Bluesky
2026-07-01 11:14:30 - Anthropic: For PRO, MAX, TEAM and select Enterprise plans, Fable 5 will be included for up to 50% of your weekly usage limit until July 7.
-
@jeremychone.bsky.social on Bluesky
#AIPack v0.8.29 released Sonnet 5 alias/pricing (and fable and mythos) z.ai & kimi providers (for GLM 5.2 and Kimi 2p7) OMLX for Mac local support Many fixes aipack.ai Built in #rustlang
-
@michaelnemtsev.bsky.social on Bluesky
Today's picks: • Claude Sonnet 5: Anthropic cuts agent pricing to $2 per million input tokens • X launches a hosted MCP server so Claude and Cursor can read the platform • Base44 trains its own model, Base1, to stop paying frontier labs per token • Google's Nano Banana 2 Lite makes images a… 2/3 ↓
-
immersive-translate/immersive-translate — ### 插件版本号 | Version 1.30.3 翻译引擎 | Translation Engine Claude Sonnet 5（内置 AI 大模型服务） 平台 | Platform macOS 浏览器 | Browsers Chrome 插件类型 | Extension Type 浏览器插件 | Browser Extension 请描述这个Bug | Describe the bug 在插件的 AI 大模型服务列表中，点击「Claude Sonnet 5」卡片的"点此测试服务"，测试失败，报错： 服务返回错误， 请求参数错误，请检查相应配置。 400: temperature is deprecated for this model. 查阅 Anthropic 官方文档确认：从 Claude Opus 4.7 / 4.8、Claude Sonnet 5 起，temperature、top_p、top_k 采样参数已被移除/受限，非默认值会直接返回 400（Fable 5 / Opus 4.7
-
GPT-5.6's biggest cost lever for builders isn't the price cut. It's how it caches. OpenAI put GPT-5.6 (Sol / Terra / Luna) into limited preview via the API and Codex, and the quiet change is prompt caching you can actually control: - explicit cache breakpoints — you decide https://t.co/lNfsFmHYPa
-
@nic221@techhub.social on Mastodon
Trump administration to lift restrictions on Anthropic's Fable 5 https://www. axios.com/2026/06/30/trump-ant hropic-ai-model-fable-restrictions # AI # Anthropic
-
We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5. We'll begin restoring access tomorrow, and will share an update soon. We’re grateful to our users for their patience, and to everyone who worked with us on
-
【JetBrains社による120億人の専門家混合モデル「Mellum2」をご紹介します。】 https:// huggingface.co/blog/JetBrains/ mellum2-launch ※AI生成の自動投稿（見出し＋リンク） # AI # 生成AI # LLM # AIGenerated
-
@engadget@robot.villas on Mastodon
Anthropic's new Sonnet 5 model is better at the tasks that are running up enterprise bills https://www.engadget.com/2205475/anthropic-releases-claude-sonnet-5-model/ # AI # Tech # LLM
-
Anthropic has announced new pricing for its Claude Sonnet 5 model. Through August 31, it costs $2 per 1M input tokens and $10 per 1M output tokens. After this date, prices will rise to $3 and $15 respectively. This update signals a shift in AI pricing strategy, potentially
-
Claude Sonnet 5 is now available in Cursor. On CursorBench, it's a meaningful step up from Sonnet 4.6: 57% vs. 49%. https://t.co/AQVHzrvqcR
-
Introducing Claude Sonnet 5, our most agentic Sonnet yet. It makes plans, uses tools like browsers and terminals, and runs autonomously at a level that just a few months ago required larger and more expensive models. https://t.co/UKK8G7ww5h
-
claude code 撞到 server 的限制... ● API Error: Server is temporarily limiting requests (not your usage limit) · Rate limited
-
Anthropic has launched Claude Science, a workbench that gives scientists one environment for computational research, saving them from bouncing between databases, pipelines and tools. The move bets on workflow over a new model to win over researchers. https:// techcrunch.com/2026/06/30/anth ropics-claude-science-bets-on-workflow-not-a-new-model-to-win-over-scientists/ # AIagent # AI # GenAI # AIResearch
-
@newsletter-tf.bsky.social on Bluesky
Replit Pivots to AI Agents, Banking on "Programming in English" Replit now uses AI agents for coding tasks, changing its pricing to pay-per-use. This affects developers and businesses using the platform. #ReplitAI, #AICoding, #Prog... https://newsletter.tf/replit-ai-agents-coding-pricing-changes/
-
@newsletter-tf.bsky.social on Bluesky
Replit is now using AI agents for coding, moving away from old methods and changing how customers pay. https://newsletter.tf/replit-ai-agents-coding-pricing-changes/
-
Introducing Claude Science, a new app designed with every stage of research in mind. Artifacts traced to their code, environments managed on demand, and 60+ optional scientific databases that you can connect. Available now in beta. https://t.co/HKhLknxLJO
-
Deprecated: no longer needed on Opus 4.8+ (silent-stop fixed model-side)
evnchn-agentic/claude-code-stop-gate — *Posted by Claude Code on evnchn's behalf.* TL;DR: Deprecated — this hook is no longer needed on Claude Opus 4.8+. The silent-stop failure this gate guarded against — the model ending a turn text-only when the task still needed a tool call — was fixed model-side in Opus 4.8, not in the Claude Code harness. The guard is now redundant on 4.8+. Keep it only if you run Opus 4.7 or older, where the failure mode returns and you'd be unguarded. Archived as a depre
-
📢 Microsoft is changing Microsoft 365 licensing tomorrow. And it's not just another price increase. Over the past 90 days, Microsoft has introduced more licensing changes than we've seen in years. 💰 New pricing. 📦 New bundles. 🤖 New AI and Copilot licensing models. 📊
-
We’re shipping 2 major releases: 🔘 Nano Banana 2 Lite: our fastest and cheapest Gemini Image model 🔘 Gemini Omni Flash: now available via the Gemini API and in @GoogleAIStudio to help developers generate and edit high-quality videos. https://t.co/fqB2sA5Xyl
-
I switched from Claude Sonnet/Opus to mostly using Kimi K2.7 Code and GLM-5.2, and... nothing bad happened. The @hf.co Inference API makes it super easy to switch between different models and providers: huggingface.co/inference/mo... Open models are really competitive - great pricing and fast.
-
@shawnchauhan1.bsky.social on Bluesky
Peak pricing on APIs tells you demand outstrips supply. DeepSeek‑V4’s mid‑July full release with peak-hour price doubling signals real capacity constraints. Plan for cost volatility and degraded UX during spikes. Consider adaptive pricing, throttles, or cheaper fallback modes.
-
Replace deprecated Groq model (llama-3.3-70b-versatile)
codedaily04/commit-craft — ## Description The project currently uses the llama-3.3-70b-versatile Groq model, which has been deprecated by Groq. According to Groq's announcement, the model will be decommissioned on August 16, 2026. After that date, API requests using this model will no longer be served. Suggested Fix Replace llama-3.3-70b-versatile with one of Groq's recommended replacement models, such as: gpt-oss-120b qwen3.6-27b It would also be beneficial to make the model name configurable t
-
DeepSeek V4 official release coming in mid-July with 2x peak-hour API pricing
-
Cursor for iOS から Pricing 見るとやたら高い気がする。Apple 税が上乗せされてる？
-
@sylphy-echosphere.bsky.social on Bluesky
でも、段を増やした瞬間に注釈も増えます。OpenAIは、複雑な推論は遅く高コストになりうると案内しているし、Gemini APIの料金表も当然トークン課金。『自然さが別物』でも、『タダで放置』にはなりません。 https://developers.openai.com/api/docs/guides/prompt-engineering https://ai.google.dev/gemini-api/docs/pricing
-
@yomimonoid.bsky.social on Bluesky
Google Makes Personalized Image Generation Free for US Gemini Users Gemini's free tier now generates images from a user's own photo library. The privacy implications of that data pipeline are the feature's real weight.
-
@aistorynews@techhub.social on Mastodon
TechCrunch reports the California Claude deal lets state agencies buy Anthropic’s AI at half price. See how this discount could reset https:// aistory.news/ai-startups-and-c ompanies/california-claude-deal-gives-state-half-price-access/ # AIHardware # Anthropic # ChatGPT
-
@joshuashew.bsky.social on Bluesky
Out of Claude usage limit for the next 3 hours, and I’m finding Minimax M3 is really struggling with the all the tooling I typically use with Claude Opus 4.8… I’ve been spoiled by regularly using a smarter model, and I really need to improve the rough edges to make it a Minimax-friendly setup.
-
@qiaokezhizao.bsky.social on Bluesky
DeepSeek V4 charging 2x for API calls during peak hours (9-12, 2-6) proves flat-rate LLM pricing is dead. If your system lacks a queue to shift batch inference to 2 AM, your costs just doubled. Classic cloud capacity planning is back, just for AI tokens.
-
@sourcenouveau.bsky.social on Bluesky
Is Anthropic really going to take back the 50% increase it gave on weekly limits? On Pro the higher limit is still very constraining. I'll be watching as July 13 approaches. xcancel.com/ClaudeDevs/s...
-
@georgesl.bsky.social on Bluesky
DeepSeek pricing changing to peak hour rates in mid-July that are 2x regular pricing. Use my timezone scheduler app to figure out peak hours for your city timezones.centminmod.com Peak hours (in UTC): 1:00–4:00 AM and 6:00–10:00 AM. (UTC+8 equivalent: 9:00 AM–12:00 noon and 2:00–6:00 PM.)
No signals match these filters.