{"exportVersion":1,"sourceSchemaVersion":1,"feedAsOf":"2026-09-30","selection":{"providerId":null,"category":"all"},"feedUrl":"https://stackreplay.com/changelog/feed.json","items":[{"url":"https://stackreplay.com/changelog/gemini-4-argon-announced","categories":["models"],"event":{"id":"gemini-4-argon-announced","type":"model_announcement","importance":"major","occurredAt":"2026-09-30","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"google","modelIds":["gemini-4-argon"],"planIds":[],"benchmark":{"sourceSetIds":["google-deepmind-argon-2026-09-30"],"benchmarkIds":[],"headlineBenchmarkId":"deep-swe-v1-1"},"title":"Gemini 4 Argon announced","summary":"Google's new frontier model is rolling out first to trusted cyber defenders. Paid API customers and Google AI Ultra subscribers come next; no API date is published yet.","status":"announced","sources":[{"url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon","title":"Gemini 4 Argon: our next era of frontier intelligence","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Today, we're announcing our new frontier model, Gemini 4 Argon, which is rolling out to a set of trusted cyber defenders through our Fairwind Program."}]}},{"url":"https://stackreplay.com/changelog/claude-sonnet-4-5-deprecated","categories":["models"],"event":{"id":"claude-sonnet-4-5-deprecated","type":"model_retirement","importance":"minor","occurredAt":"2026-09-30","dateBasis":"provider_publication_date","effectiveAt":"2026-11-30","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"anthropic","modelIds":["claude-sonnet-4-5"],"planIds":[],"title":"Claude Sonnet 4.5 deprecated","summary":"Anthropic notified API developers that Claude Sonnet 4.5 retires on November 30, 2026, with Claude Sonnet 5.5 as the recommended replacement.","status":"scheduled","sources":[{"url":"https://platform.claude.com/docs/en/about-claude/model-deprecations","title":"Model deprecations - Claude Platform Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"2026-09-30: Claude Sonnet 4.5 model On September 30, 2026, Anthropic notified developers using Claude Sonnet 4.5 of its upcoming retirement on the Claude API. Retirement date November 30, 2026 claude-sonnet-4-5-20250929 claude-sonnet-5-5"}]}},{"url":"https://stackreplay.com/changelog/google-argon-launch-results","categories":["benchmarks"],"event":{"id":"google-argon-launch-results","type":"benchmark_update","importance":"minor","occurredAt":"2026-09-30","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"google","modelIds":["gemini-4-argon","gpt-6-astra","claude-fable-5-1","claude-opus-5-5"],"planIds":[],"benchmark":{"sourceSetIds":["google-deepmind-argon-2026-09-30"],"benchmarkIds":[],"headlineBenchmarkId":"deep-swe-v1-1"},"title":"Google publishes Gemini 4 Argon results on 17 benchmarks","summary":"Google's launch table compares Gemini 4 Argon with GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5 across coding, knowledge work, science, long context, multimodal and security benchmarks.","status":"effective","sources":[{"url":"https://deepmind.google/models/gemini/","title":"Gemini 4 Argon (Google DeepMind)","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Benchmark Notes Gemini 4 Argon GPT-6 Astra Claude Fable 5.1 Claude Opus 5.5 Vals Index Knowledge work 68.9% 63.1% 65.8% 67.0%"}]}},{"url":"https://stackreplay.com/changelog/chatgpt-pro-500-launched","categories":["subscriptions"],"event":{"id":"chatgpt-pro-500-launched","type":"plan_launch","importance":"major","occurredAt":"2026-09-29","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":[],"planIds":["openai-chatgpt-pro-500"],"title":"ChatGPT Pro 500 launched at $500 a month","summary":"A third Pro tier with the highest included usage of the Pro plans and Astra Ultrafast in ChatGPT Work and Codex. Pro 100 and Pro 200 remain.","status":"available","sources":[{"url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes","title":"ChatGPT - Release Notes","authority":"first_party","checkedAt":"2026-09-30","excerpt":"We're introducing Pro 500, a new $500/month plan with the highest included usage of the Pro plans and access to Astra Ultrafast in ChatGPT Work and Codex."},{"url":"https://help.openai.com/en/articles/9793128-about-chatgpt-pro-tiers","title":"About ChatGPT Pro tiers","authority":"first_party","checkedAt":"2026-09-30","excerpt":"ChatGPT Pro now offers Pro 500, a new $500/month plan that includes Astra Ultrafast. Pro 200 is also available for new subscriptions again. New subscriptions that aren't eligible for grandfathering include a lower usage allowance."}]}},{"url":"https://stackreplay.com/changelog/gpt-6-1-sol-released","categories":["models"],"event":{"id":"gpt-6-1-sol-released","type":"model_release","importance":"major","occurredAt":"2026-09-29","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":["gpt-6-1-sol"],"planIds":[],"benchmark":{"sourceSetIds":["openai-sol-2026-09-29-max"],"benchmarkIds":[],"headlineBenchmarkId":"deep-swe-v1-1"},"title":"GPT-6.1 Sol released","summary":"OpenAI's new Sol model for complex coding and professional work, at a lower cost than GPT-6 Astra. Available in the API and in ChatGPT Work and Codex on paid plans.","status":"available","sources":[{"url":"https://developers.openai.com/api/docs/changelog","title":"Changelog | OpenAI API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sep 29 Feature gpt-6.1-sol v1/responses v1/chat/completions Released GPT-6.1 Sol ( gpt-6.1-sol ) for complex coding and professional work at a lower cost than GPT-6 Astra."},{"url":"https://openai.com/index/introducing-gpt-6-1-sol/","title":"Introducing GPT-6.1 Sol","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Its standard API prices are $2 per million input tokens, $0.10 per million cached input tokens, and $10 per million output tokens."},{"url":"https://developers.openai.com/api/docs/models/gpt-6.1-sol","title":"GPT-6.1 Sol Model | OpenAI API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"1,050,000 context window 128,000 max output tokens"}]}},{"url":"https://stackreplay.com/changelog/chatgpt-pro-200-reopened-lower-allowance","categories":["subscriptions"],"event":{"id":"chatgpt-pro-200-reopened-lower-allowance","type":"plan_limit_change","importance":"notable","occurredAt":"2026-09-29","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":[],"planIds":["openai-chatgpt-pro-20x"],"title":"ChatGPT Pro 200 reopens with a lower usage allowance","summary":"New subscribers can join Pro 200 again at $200 a month, with less included usage than before. Subscribers eligible for grandfathering keep the previous allowance through October 29.","status":"effective","sources":[{"url":"https://help.openai.com/en/articles/9793128-about-chatgpt-pro-tiers","title":"About ChatGPT Pro tiers","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Yes. New Pro 200 subscriptions are available from the pricing page. New subscriptions that aren't eligible for grandfathering include a lower usage allowance than previously offered with Pro 200 to reflect our increasingly efficient models. The monthly price remains $200."},{"url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes","title":"ChatGPT - Release Notes","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Pro 200 subscriptions: Pro 200 is available for new subscriptions again at $200/month. Included usage allowances are changing."}]}},{"url":"https://stackreplay.com/changelog/gpt-6-astra-ultrafast","categories":["models"],"event":{"id":"gpt-6-astra-ultrafast","type":"api_availability","importance":"notable","occurredAt":"2026-09-29","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":["gpt-6-astra"],"planIds":["openai-chatgpt-pro-500"],"title":"GPT-6 Astra Ultrafast added to the API","summary":"A faster service tier for GPT-6 Astra: up to 6x faster generation in the API and up to 8x in Codex. In ChatGPT it is limited to Pro 500 and Enterprise.","status":"available","sources":[{"url":"https://developers.openai.com/api/docs/changelog","title":"Changelog | OpenAI API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sep 29 Feature gpt-6-astra v1/responses Added Ultrafast mode for GPT-6 Astra in the Responses API."},{"url":"https://openai.com/index/devday-2026-recap/","title":"DevDay 2026 Recap","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Ultrafast offers up to 8x faster token generation (300 tokens per second) in Codex and up to 6x in the API."}]}},{"url":"https://stackreplay.com/changelog/claude-sonnet-5-5-released","categories":["models"],"event":{"id":"claude-sonnet-5-5-released","type":"model_release","importance":"major","occurredAt":"2026-09-28","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"anthropic","modelIds":["claude-sonnet-5-5"],"planIds":[],"benchmark":{"sourceSetIds":["anthropic-sonnet-2026-09-28"],"benchmarkIds":[],"headlineBenchmarkId":"terminal-bench-4-0"},"title":"Claude Sonnet 5.5 released","summary":"Anthropic's new mid-tier model, priced the same as Sonnet 5 and available on the Claude API, Amazon Bedrock, Google Cloud and Microsoft Foundry.","status":"available","sources":[{"url":"https://www.anthropic.com/claude-sonnet-5-5","title":"Introducing Claude Sonnet 5.5","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sonnet 5.5 is priced the same as Sonnet 5 at $2 per million input tokens, $10 per million output tokens, and $0.20 per million tokens for cache reads, but it typically needs far fewer tokens to do the same work."},{"url":"https://platform.claude.com/docs/en/about-claude/model-deprecations","title":"Model deprecations - Claude Platform Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"claude-sonnet-5-5 Active N/A Not sooner than September 28, 2027"}]}},{"url":"https://stackreplay.com/changelog/anthropic-sonnet-5-5-results","categories":["benchmarks"],"event":{"id":"anthropic-sonnet-5-5-results","type":"benchmark_update","importance":"minor","occurredAt":"2026-09-28","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"anthropic","modelIds":["claude-sonnet-5-5"],"planIds":[],"benchmark":{"sourceSetIds":["anthropic-sonnet-2026-09-28"],"benchmarkIds":[],"headlineBenchmarkId":"terminal-bench-4-0"},"title":"Anthropic publishes Claude Sonnet 5.5 launch results","summary":"Anthropic's launch results for Sonnet 5.5 include Terminal-Bench 4.0, where it reports 70.6% against Sonnet 5's 10.3%.","status":"effective","sources":[{"url":"https://www.anthropic.com/claude-sonnet-5-5","title":"Introducing Claude Sonnet 5.5","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sonnet 5.5 scores 70.6% on Terminal-Bench 4.0, an agentic coding evaluation, compared to Sonnet 5’s 10.3%."}]}},{"url":"https://stackreplay.com/changelog/claude-opus-5-5-released","categories":["models"],"event":{"id":"claude-opus-5-5-released","type":"model_release","importance":"major","occurredAt":"2026-09-22","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"anthropic","modelIds":["claude-opus-5-5"],"planIds":[],"title":"Claude Opus 5.5 released","summary":"Anthropic's new flagship. Input and output cost 20% less than Opus 5, and cache reads, most of the cost of agentic coding, 60% less.","status":"available","sources":[{"url":"https://www.anthropic.com/claude-opus-5-5","title":"Introducing Claude Opus 5.5","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Input and output tokens are $4 and $20 per million, 20% less than Opus 5. Cache reads (which make up the majority of agentic and coding work costs) are $0.20 per million tokens, 60% less than Opus 5."},{"url":"https://platform.claude.com/docs/en/release-notes/api","title":"Claude API Release Notes","authority":"first_party","checkedAt":"2026-09-30","excerpt":"September 22, 2026 ... On Claude Opus 5.5, thinking can't be disabled: thinking: {\"type\": \"disabled\"} and thinking: {\"type\": \"enabled\", ...} return a 400 error."}]}},{"url":"https://stackreplay.com/changelog/gpt-6-sol-luna-released","categories":["models"],"event":{"id":"gpt-6-sol-luna-released","type":"model_release","importance":"major","occurredAt":"2026-09-22","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":["gpt-6-sol","gpt-6-luna"],"planIds":[],"title":"GPT-6 Sol and GPT-6 Luna released","summary":"OpenAI's new Sol and Luna models, with API prices 50% below GPT-5.6 Sol and Luna's promotional pricing.","status":"available","sources":[{"url":"https://developers.openai.com/api/docs/changelog","title":"Changelog | OpenAI API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sep 22 Feature gpt-6-sol gpt-6-luna v1/responses v1/chat/completions Released GPT-6 Sol ( gpt-6-sol ) and GPT-6 Luna ( gpt-6-luna )."},{"url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","title":"Introducing GPT-6 Sol and Luna","authority":"first_party","checkedAt":"2026-09-30","excerpt":"GPT-5.6 Sol to GPT-6 Sol $4 to $2 / $20 to $10 50% cheaper. GPT-5.6 Luna to GPT-6 Luna $0.20 to $0.10 / $1.20 to $0.50 50% cheaper. Prices are per 1 million tokens."}]}},{"url":"https://stackreplay.com/changelog/claude-five-hour-limits-raised","categories":["subscriptions"],"event":{"id":"claude-five-hour-limits-raised","type":"plan_limit_change","importance":"notable","occurredAt":"2026-09-22","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"anthropic","modelIds":[],"planIds":["anthropic-claude-pro","anthropic-claude-max-5x","anthropic-claude-max-20x"],"title":"Claude raises five-hour usage limits on paid plans","summary":"Alongside Opus 5.5, Anthropic increased five-hour usage limits on Pro, Max, Team and seat-based Enterprise plans, and gave subscribers a rate-limit reset they can save and use later.","status":"effective","sources":[{"url":"https://www.anthropic.com/claude-opus-5-5","title":"Introducing Claude Opus 5.5","authority":"first_party","checkedAt":"2026-09-30","excerpt":"In addition to the price drop, we're increasing five-hour usage limits on Pro, Max, Team, and seat-based Enterprise plans. We're also providing subscription users a rate limit reset, which you can now save and use whenever you choose."}]}},{"url":"https://stackreplay.com/changelog/xiaomi-mimo-v2-6-released","categories":["models"],"event":{"id":"xiaomi-mimo-v2-6-released","type":"model_release","importance":"minor","occurredAt":"2026-09-22","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"xiaomi","modelIds":["mimo-v2-6-pro","mimo-v2-6-flash"],"planIds":[],"title":"Xiaomi MiMo-V2.6 Pro and Flash released with open weights","summary":"Xiaomi released and open-sourced the MiMo-V2.6 series, two multimodal models at the same API pricing as V2.5.","status":"available","sources":[{"url":"https://mimo.mi.com/docs/en-US/news/latest/v2-6","title":"MiMo-V2.6: Scaling Up Reinforcement Learning for Self-Improvement | Xiaomi MiMo","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Today, we are officially releasing and open-sourcing the Xiaomi MiMo-V2.6 series. ... The MiMo-V2.6 series adopts the same API pricing as the V2.5 series."},{"url":"https://mimo.mi.com/models/en-US/mimo-v2.6-flash","title":"MiMo-V2.6-Flash | Xiaomi MiMo","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Input (cache miss) $0.14 USD / Million tokens ... Output $0.28 USD / Million tokens ... Context Window 1M tokens"}]}},{"url":"https://stackreplay.com/changelog/grok-4-7-released","categories":["models"],"event":{"id":"grok-4-7-released","type":"model_release","importance":"major","occurredAt":"2026-09-21","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"x-ai","modelIds":["grok-4-7"],"planIds":[],"benchmark":{"sourceSetIds":["spacexai-grok-2026-09-21"],"benchmarkIds":[],"headlineBenchmarkId":"deep-swe-v1-1"},"title":"Grok 4.7 released","summary":"xAI's new frontier model for coding, agentic tasks and knowledge work, now on the xAI API with a 500K context window.","status":"available","sources":[{"url":"https://docs.x.ai/developers/release-notes","title":"Release Notes | xAI Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"September 21 Grok 4.7 Grok 4.7, SpaceXAI's frontier model for coding, agentic tasks, and knowledge work, is now available on the xAI API as grok-4.7 . It has a 500k context window, text and image inputs with text-only output, and no text output limit. Pricing is $2 / $0.50 / $6 per 1M tokens"},{"url":"https://x.ai/news/grok-4-7","title":"Introducing Grok 4.7","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sep 21, 2026 ... The model is priced starting at $2 per million input tokens and $6 per million output tokens. We also serve a fast variant with twice the output speed at twice the price."}]}},{"url":"https://stackreplay.com/changelog/xai-grok-4-7-results","categories":["benchmarks"],"event":{"id":"xai-grok-4-7-results","type":"benchmark_update","importance":"minor","occurredAt":"2026-09-21","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"x-ai","modelIds":["grok-4-7"],"planIds":[],"benchmark":{"sourceSetIds":["spacexai-grok-2026-09-21"],"benchmarkIds":[],"headlineBenchmarkId":"deep-swe-v1-1"},"title":"xAI publishes Grok 4.7 launch results","summary":"xAI's launch table reports Grok 4.7 on CursorBench 4.0, DeepSWE v1.1, EEBench, Terminal-Bench 4.0 and HealthBench Professional, among others.","status":"effective","sources":[{"url":"https://x.ai/news/grok-4-7","title":"Introducing Grok 4.7","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Token prices and benchmark scores for Grok 4.7, Grok 4.6, GPT-5.6 Sol, and Fable 5.1."}]}},{"url":"https://stackreplay.com/changelog/qwen-3-8-omni-flash-released","categories":["models"],"event":{"id":"qwen-3-8-omni-flash-released","type":"model_release","importance":"notable","occurredAt":"2026-09-18","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"alibaba","modelIds":["qwen-3-8-omni-flash"],"planIds":[],"title":"Qwen3.8-Omni-Flash released","summary":"Qwen's next native omni-modal model takes text, image, audio and video input with up to 1M tokens of context, on Alibaba Cloud Model Studio.","status":"available","sources":[{"url":"https://docs.qwencloud.com/changelog","title":"Changelog | QwenCloud Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"September 18, 2026 ... qwen3.8-omni-flash Qwen's next-generation native omni-modal model supports context lengths of up to 1M tokens and natively accepts text, image, audio, and video inputs."}]}},{"url":"https://stackreplay.com/changelog/deepseek-v4-1-flash-released","categories":["models"],"event":{"id":"deepseek-v4-1-flash-released","type":"model_release","importance":"major","occurredAt":"2026-09-10","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"deepseek","modelIds":["deepseek-v4-1-flash"],"planIds":[],"benchmark":{"sourceSetIds":["deepseek-flash-2026-09-10"],"benchmarkIds":[],"headlineBenchmarkId":"deep-swe-v1-1"},"title":"DeepSeek-V4.1-Flash released","summary":"DeepSeek's first model in a new architecture family, with native visual understanding, a 1M-token context and open weights.","status":"available","sources":[{"url":"https://api-docs.deepseek.com/updates/","title":"Change Log | DeepSeek API Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Date: 2026-09-10 DeepSeek-V4.1-Flash Release Today, we officially release the DeepSeek-V4.1-Flash model. It is the smallest model in our new architecture family, with native multimodal visual understanding."},{"url":"https://www.deepseek.com/en/news/deepseek-v4-1-flash/","title":"Introducing DeepSeek-V4.1-Flash","authority":"first_party","checkedAt":"2026-09-30","excerpt":"552B-parameter MoE. New Causal Encoder-Decoder architecture: just 8B active parameters for input, 16B for output."}]}},{"url":"https://stackreplay.com/changelog/chatgpt-pro-200-paused","categories":["subscriptions"],"event":{"id":"chatgpt-pro-200-paused","type":"plan_availability_change","importance":"notable","occurredAt":"2026-09-10","dateBasis":"provider_stated_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":[],"planIds":["openai-chatgpt-pro-20x"],"title":"ChatGPT Pro 200 pauses new sign-ups","summary":"OpenAI paused new sign-ups and upgrades to the $200 Pro plan. Existing Pro 200 and Pro 100 subscriptions were not affected.","status":"paused","sources":[{"url":"https://web.archive.org/web/20260919102801/https://help.openai.com/en/articles/9793128-about-chatgpt-pro-tiers","title":"About ChatGPT Pro tiers (OpenAI Help Center), archived Sep 19, 2026","authority":"first_party","checkedAt":"2026-09-29","excerpt":"As of September 10, 2026, we're temporarily pausing new sign-ups and upgrades to the ChatGPT Pro $200 plan (Pro 20X)."}]}},{"url":"https://stackreplay.com/changelog/deepseek-api-prices-cut","categories":["pricing"],"event":{"id":"deepseek-api-prices-cut","type":"api_price_change","importance":"notable","occurredAt":"2026-09-10","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"deepseek","modelIds":["deepseek-v4-1-flash"],"planIds":[],"title":"DeepSeek lowers API prices with V4.1-Flash","summary":"DeepSeek cut its API prices with the V4.1-Flash release, effective 04:00 UTC on September 10. Off-peak rates stay at half the peak rate.","status":"effective","sources":[{"url":"https://www.deepseek.com/en/news/deepseek-v4-1-flash/","title":"Introducing DeepSeek-V4.1-Flash","authority":"first_party","checkedAt":"2026-09-30","excerpt":"More efficient architecture. Lower API prices. ... New pricing takes effect at 04:00 UTC on Sept 10, 2026."},{"url":"https://api-docs.deepseek.com/updates/","title":"Change Log | DeepSeek API Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"API pricing adjustment With the release of DeepSeek-V4.1-Flash, API prices have been reduced accordingly."},{"url":"https://api-docs.deepseek.com/quick_start/pricing","title":"Models & Pricing | DeepSeek API Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"PRICING (2) 1M INPUT TOKENS (CACHE MISS) OFF-PEAK $0.15 ... PEAK $0.3 ... 1M OUTPUT TOKENS OFF-PEAK $0.6 ... PEAK $1.2"}]}},{"url":"https://stackreplay.com/changelog/deepseek-v4-flash-retired","categories":["models"],"event":{"id":"deepseek-v4-flash-retired","type":"model_retirement","importance":"minor","occurredAt":"2026-09-10","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"deepseek","modelIds":["deepseek-v4-flash","deepseek-v4-flash-vision-exp"],"planIds":[],"title":"DeepSeek V4 Flash and V4 Flash Vision Exp retired","summary":"Both previous-generation models were retired. Their model names temporarily route to V4.1-Flash.","status":"retired","sources":[{"url":"https://api-docs.deepseek.com/updates/","title":"Change Log | DeepSeek API Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"The previous-generation models V4 Flash and V4 Flash Vision Exp have been retired; for compatibility, the model names deepseek-v4-flash and deepseek-v4-flash-vision-exp are temporarily routed to V4.1 Flash."}]}},{"url":"https://stackreplay.com/changelog/gpt-6-astra-released","categories":["models"],"event":{"id":"gpt-6-astra-released","type":"model_release","importance":"major","occurredAt":"2026-09-03","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"openai","modelIds":["gpt-6-astra"],"planIds":[],"title":"GPT-6 Astra released","summary":"OpenAI's most capable model, built for the hardest end-to-end work, released in the API. The ChatGPT rollout began with a limited set of organizations.","status":"available","sources":[{"url":"https://developers.openai.com/api/docs/changelog","title":"Changelog | OpenAI API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Sep 3 Feature gpt-6-astra v1/responses v1/chat/completions Released GPT-6 Astra , our most capable model, built for the hardest end-to-end work."},{"url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes","title":"ChatGPT - Release Notes","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Today we're introducing GPT-6 Astra, with improvements in coding, research, computer use, and complex, multi-step work."},{"url":"https://developers.openai.com/api/docs/models/gpt-6-astra","title":"GPT-6 Astra Model | OpenAI API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Input $10.00 Cached input $1.00 Cache writes $12.50 Output $50.00"}]}},{"url":"https://stackreplay.com/changelog/gemini-3-8-flash-released","categories":["models"],"event":{"id":"gemini-3-8-flash-released","type":"model_release","importance":"notable","occurredAt":"2026-09-02","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"google","modelIds":["gemini-3-8-flash"],"planIds":[],"title":"Gemini 3.8 Flash generally available","summary":"Google's most capable Flash model, aimed at long-horizon coding, agents and enterprise workflows, at the same introductory price as 3.7 Flash through December 31.","status":"available","sources":[{"url":"https://ai.google.dev/gemini-api/docs/changelog?hl=en","title":"Release notes | Gemini API","authority":"first_party","checkedAt":"2026-09-30","excerpt":"September 2, 2026 Gemini 3.8 Flash generally available (GA) : Released gemini-3.8-flash , our most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows."},{"url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/","title":"Introducing Gemini 3.8 Flash and 3.8 Flash Cyber","authority":"first_party","checkedAt":"2026-09-30","excerpt":"It is available at the same introductory price as 3.7 Flash at $0.75 per million input tokens and $3.75 per million output tokens. ... Introductory price expires on December 31, 2026. Starting January 1, 2027, $1.50/1M input tokens and $7.50/1M output tokens will apply."}]}},{"url":"https://stackreplay.com/changelog/muse-spark-1-3-released","categories":["models"],"event":{"id":"muse-spark-1-3-released","type":"model_release","importance":"notable","occurredAt":"2026-09-02","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"meta","modelIds":["muse-spark-1-3"],"planIds":[],"title":"Meta Muse Spark 1.3 released","summary":"Meta's updated Muse Spark model, available with max reasoning in Muse Code and the Meta Model API.","status":"available","sources":[{"url":"https://research.meta.ai/blog/introducing-muse-spark-1-3","title":"Introducing Muse Spark 1.3 | Meta AI Research","authority":"first_party","checkedAt":"2026-09-30","excerpt":"September 2, 2026 - 3 minute read ... Muse Spark 1.3 with max reasoning is now available on Muse Code and Meta Model API. Get started at dev.meta.ai ."}]}},{"url":"https://stackreplay.com/changelog/qwen-3-8-max-0902-snapshot","categories":["models"],"event":{"id":"qwen-3-8-max-0902-snapshot","type":"model_release","importance":"minor","occurredAt":"2026-09-02","dateBasis":"provider_publication_date","effectiveAt":"2026-09-05","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"alibaba","modelIds":["qwen-3-8-max-0902"],"planIds":[],"title":"Qwen3.8-Max 0902 snapshot released","summary":"An upgraded Qwen3.8-Max snapshot. The qwen3.8-max endpoint moved to it on September 5 with pricing unchanged.","status":"effective","sources":[{"url":"https://docs.qwencloud.com/changelog","title":"Changelog | QwenCloud Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"September 2, 2026 qwen3.8-max-0902 ... is an upgraded snapshot of qwen3.8-max."},{"url":"https://docs.qwencloud.com/changelog/model-updates/qwen3.8-max-upgrade","title":"Qwen3.8-Max Update Notice - QwenCloud","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Effective September 5, 2026 (UTC+8), QwenCloud will upgrade the qwen3.8-max model. After the upgrade, the qwen3.8-max endpoint will automatically transition to the snapshot version qwen3.8-max-0902, with billing items and pricing remaining unchanged."}]}},{"url":"https://stackreplay.com/changelog/claude-fable-5-1-released","categories":["models"],"event":{"id":"claude-fable-5-1-released","type":"model_release","importance":"notable","occurredAt":"2026-09-01","dateBasis":"provider_publication_date","discoveredAt":"2026-09-30","verifiedAt":"2026-09-30","providerId":"anthropic","modelIds":["claude-fable-5-1"],"planIds":[],"title":"Claude Fable 5.1 released","summary":"A new Claude Fable release with a 1M-token context window and 128K maximum output.","status":"available","sources":[{"url":"https://platform.claude.com/docs/en/models/fable-5-1/overview","title":"Claude Fable 5.1 - Claude Platform Docs","authority":"first_party","checkedAt":"2026-09-30","excerpt":"Claude Fable 5.1 This model 1M 128K $10 / $50 Slower Adaptive (always on) high Jun 2026 ... Released September 1, 2026 Retirement Not sooner than September 1, 2027"}]}}]}