Changelog
2026-10-14
Changed
- Restructured Alibaba Token Plan. Now Personal plans at $6, $10, $18, and $68 monthly, plus Team seats.
- Restructured Aoru plans. Now Plus $12 and Pro $25 monthly.
- Restructured Entrim plans. Qwen 3.6 and DeepSeek V4 Flash plans retired.
- Restructured Solheim to model, instance, and context pricing. Starts at €15 monthly.
- Restructured Ollama Cloud plans. Now $20, $100, and $500 monthly, with usage credits.
- Updated HyperCLI plans. Added Team $79 and Pro $149.
- Updated CheapestInference plans. Now from $22, $71, and $199 monthly.
- Updated Yolo-Auto plans. Now Builder $19, Pro $39, and Fleet $99.
- Updated Arli plans. Added Starter $10 and Core $15.
- Updated Canopy Wave plans. MiniMax M3 now included.
- Added Go Plus at $40 to OpenCode.
- Added Max $200 to Neuralwatt Cloud.
- Added Enterprise $500 to Atlas Cloud.
- Moved StepFun Step Plan to deprecated because all tiers are out of stock.
- Moved Pareto Inference to deprecated because the subscription was removed.
2026-10-13
Changed
- Restructured xAI plans. Now Plus $8, Premium $30, Super $100, and Ultra $200. Grok 4.6 flagship at coding index 76.8.
- Updated OpenAI Codex plans. Added Pro 500 with Ultrafast for GPT-6 Astra and GPT-6.1 Sol.
- Restructured Kimi plans. Now Plus $19, Pro $39, Max $99, and Ultra $199 monthly, with annual discounts.
- Renamed MiniMax Token Plan to M Plan. Now Go, Explore, and Build yearly tiers.
- Restructured QwenCloud plans. Added Essential tier and lower prices.
- Updated Meta Muse Code. Added Maximum Usage at $200. Muse Spark 1.3 flagship at coding index 76.5.
- Updated Google plans. Added Gemini 4 Argon to Pro and Ultra.
- Updated Anthropic plans. Max tiers now include monthly API credits.
- Updated Mistral Pro. Now includes $25.5/mo in API credits. Restored Team plan.
- Updated Z.AI. GLM-5.3 now on all tiers; yearly billing is 30% off.
- Updated StepFun. All tiers currently out of stock.
- Updated Xiaomi MiMo. Now MiMo-V2.6 series.
- Updated BytePlus. Added GLM-5.3-Flash.
- Moved StreamLake KwaiKAT Coding Plan to deprecated because the plan was discontinued.
2026-09-06
Changed
- Added GPT-6 Astra access to ChatGPT Plus and Pro plans.
2026-09-04
Added
Changed
- Updated OpenCode Go. Grok 4.6 flagship at coding index 76.8.
2026-09-03
Changed
- Updated Google and Antigravity to Gemini 3.8 Flash. Coding index 76.3.
- Added Google AI Ultra 20× to Antigravity.
2026-09-02
Added
- Added Pareto Inference.
- Added Charm Hyper.
- Added HyperCLI.
2026-09-01
Added
Changed
- Updated Anthropic flagship to Claude Fable 5.1. Coding index 81.6.
- Updated LLM Gateway, ZenMux, Cursor, GitHub Copilot, Command Code, Windsurf, Factory, Warp, Zed, and VULK to Claude Fable 5.1.
2026-08-31
Added
Changed
- Removed team, business, and enterprise plans.
- Removed Augment Code because it has no qualifying individual plan.
- Removed Tabnine and Base44 because they require annual billing.
- Removed Routera because its pricing was not credible enough for inclusion.
2026-08-30
Added
2026-08-29
Added
- Added FEIHOA, an uncensored unlimited inference host.
- Added SuperGrok Plus plan at $100/mo.
- Added Velqa, a prepaid multi-model token plan.
- Added Aoru AI, a zero-log open-model host.
- Added QwenCloud Token Plan.
- Added Tencent Cloud TokenHub.
Changed
- Updated Router9 flagship to Claude Opus 5. Coding index 78.
- Updated Z.AI models. Now includes GLM-5.3-Flash.
- Updated Claudin.io. Now Essential, Pro, and Ultra tiers with Claudius Beta.
- Raised MiniMax Token Plan prices. Now $22, $55, and $132 monthly.
2026-08-28
Added
- Added Solheim, an EU-hosted unlimited endpoint provider.
Changed
- Changed StepFun quotas to monthly credits.
- Updated Yolo-Auto to Qwen3.8-27B.
2026-08-18
Changed
- Updated Z.AI flagship to GLM-5.3. Coding index now 74.8.
- Raised Z.AI Pro price to $80/mo.
- Raised Z.AI Max price to $168/mo.
2026-08-13
Added
- Added Entrim (OpenClaw Plans) as an aggregator.
- Added a changelog page.
Changed
- Updated DeepSeek V4 Pro to checkpoint 0813. Coding index now 68.8.
- Updated Artificial Analysis flagship scores. Used highest-scoring models.
