What changed in AI models, prices and terms
90 changes confirmed in the 7 days to (63 price changes, 21 listed value changes, 6 terms changes, 20 added to tracking). Last checked . Methodology · Source
Subscription limits · Terms tracker · Model prices · What's new
What changed today
No changes detected today (, our latest check). This covers only the items and values we track.
Last 7 days
- Price change: Z.ai: GLM 5.3 Flash — output price $0.50 per 1M tokens → $0.08 per 1M tokens ·
- Price change: Z.ai: GLM 5.3 Flash — input price $0.15 per 1M tokens → $0.03 per 1M tokens ·
- Price change: Qwen: Qwen3.6 27B — output price $2.70 per 1M tokens → $3.20 per 1M tokens ·
- Price change: Qwen: Qwen3.6 27B — input price $0.45 per 1M tokens → $0.32 per 1M tokens ·
- Price change: MoonshotAI: Kimi K3 (batch) (Moonshot AI) — output price $11.40 per 1M tokens → $6.00 per 1M tokens ·
- Price change: MoonshotAI: Kimi K2.5 (Moonshot AI) — input price $0.50 per 1M tokens → $0.53 per 1M tokens ·
- Price change: MoonshotAI: Kimi K2.5 (Moonshot AI) — output price $2.50 per 1M tokens → $3.33 per 1M tokens ·
- Price change: MoonshotAI: Kimi K2.5 (Moonshot AI) — $0.15 per 1M tokens → $0.21 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash (batch) — output price $0.34 per 1M tokens → $0.20 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash (batch) — $0.0034 per 1M tokens → $0.0020 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash (batch) — input price $0.11 per 1M tokens → $0.01 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash 1017 — output price $1.20 per 1M tokens → $0.50 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash 1017 — input price $0.30 per 1M tokens → $0.0060 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash 1017 — $0.0060 per 1M tokens → $0.0050 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Pro 0813 — input price $0.12 per 1M tokens → $0.11 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Pro 0813 — output price $4.00 per 1M tokens → $4.44 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash 0731 — input price $0.06 per 1M tokens → $0.04 per 1M tokens ·
- Listed value change: Qwen: Qwen3.6 27B — 65,536 tokens → 81,920 tokens ·
- Listed value change: MoonshotAI: Kimi K3 (batch) (Moonshot AI) — 16,384 tokens → 943,718 tokens ·
- Listed value change: MoonshotAI: Kimi K2.5 (Moonshot AI) — 235,929 tokens → 65,536 tokens ·
- Listed value change: DeepSeek: DeepSeek V4.1 Flash (batch) — 131,072 tokens → 943,718 tokens ·
- Listed value change: DeepSeek: DeepSeek V4.1 Flash 1017 — 393,216 tokens → 943,718 tokens ·
- Added to tracking: Z.ai: GLM 5.2 (batch). First check: $0.37 per 1M tokens ·
- Added to tracking: Morph: Morph V3 Large (batch). First check: $0.45 per 1M tokens ·
- Added to tracking: Morph: Morph V3 Fast (batch). First check: $0.40 per 1M tokens ·
- Added to tracking: DeepSeek: DeepSeek V4.1 Flash 1017 (batch). First check: $0.01 per 1M tokens ·
- Terms answer changed: Microsoft Azure OpenAI (Foundry Models) data, privacy and security — When was the Microsoft Azure OpenAI (Foundry Models) data, privacy and security last updated? 2026-05-19 → 2026-06-05 ·
- Price change: NVIDIA: Nemotron 3 Super — output price $2.40 per 1M tokens → $0.40 per 1M tokens ·
- Price change: NVIDIA: Nemotron 3 Super — input price $0.30 per 1M tokens → $0.08 per 1M tokens ·
- Price change: MiniMax: MiniMax M2.5 — $0.03 per 1M tokens → $0.06 per 1M tokens ·
- Price change: MiniMax: MiniMax M2.5 — output price $0.95 per 1M tokens → $1.20 per 1M tokens ·
- Price change: MiniMax: MiniMax M2.5 — input price $0.27 per 1M tokens → $0.30 per 1M tokens ·
- Price change: inclusionAI: Ling 3.1 Flash — input price $0.08 per 1M tokens → $0.05 per 1M tokens ·
- Price change: inclusionAI: Ling 3.1 Flash — $0.02 per 1M tokens → $0.0090 per 1M tokens ·
- Price change: inclusionAI: Ling 3.1 Flash — output price $0.23 per 1M tokens → $0.14 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash 0910 — input price $0.05 per 1M tokens → $0.05 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash 0910 — output price $0.50 per 1M tokens → $0.57 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V3 — input price $0.26 per 1M tokens → $0.32 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V3 — output price $1.03 per 1M tokens → $0.89 per 1M tokens ·
- Listed value change: NVIDIA: Nemotron 3 Super — 262,144 tokens → 16,384 tokens ·
- Listed value change: MiniMax: MiniMax M2.5 — 32,768 tokens → 58,982 tokens ·
- Listed value change: DeepSeek: DeepSeek V3 — 16,000 tokens → 16,384 tokens ·
- Added to tracking: Thinking Machines: Inkling Small (batch) (thinkingmachines). First check: $0.36 per 1M tokens ·
- Added to tracking: Ricursive: Ripple 65B Preview (ricursiveai). First check: $0.20 per 1M tokens ·
- Added to tracking: Qwen: Qwen3.8 27B (batch). First check: $0.16 per 1M tokens ·
- Added to tracking: MiniMax: MiniMax M3 (batch). First check: $0.22 per 1M tokens ·
- Added to tracking: Meta: Muse Glimmer 30B (batch). First check: $0.24 per 1M tokens ·
- Added to tracking: LiquidAI: LFM3-1.7B (liquid). First check: $0.08 per 1M tokens ·
- Added to tracking: LiquidAI: LFM3-1.7B (free) (liquid). First check: $0.00 per 1M tokens (free) ·
- Added to tracking: DeepSeek: DeepSeek V4 Flash 0731 (batch). First check: $0.05 per 1M tokens ·
- Terms answer changed: AWS Service Terms — When was the AWS Service Terms last updated? 2026-11-03 → 2026-11-05 ·
- Price change: Reflection: Beam 1.0 501B A23B — $0.05 per 1M tokens → $0.04 per 1M tokens ·
- Price change: Reflection: Beam 1.0 501B A23B — input price $0.25 per 1M tokens → $0.30 per 1M tokens ·
- Price change: MoonshotAI: Kimi K2.7 Code (Moonshot AI) — $0.15 per 1M tokens → $0.16 per 1M tokens ·
- Price change: MiniMax: MiniMax M2.7 — output price $1.20 per 1M tokens → $0.84 per 1M tokens ·
- Price change: MiniMax: MiniMax M2.7 — input price $0.30 per 1M tokens → $0.21 per 1M tokens ·
- Price change: MiniMax: MiniMax M2.7 — $0.06 per 1M tokens → $0.04 per 1M tokens ·
- Price change: Meta: Llama 3.3 70B Instruct — output price $0.32 per 1M tokens → $0.79 per 1M tokens ·
- Price change: Meta: Llama 3.3 70B Instruct — input price $0.10 per 1M tokens → $0.59 per 1M tokens ·
- Price change: Meta: Llama 3.3 70B Instruct — $0.10 per 1M tokens → $0.30 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash (batch) — $0.0040 per 1M tokens → $0.0034 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash (batch) — input price $0.05 per 1M tokens → $0.11 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash (batch) — output price $0.15 per 1M tokens → $0.34 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash Vision Exp — input price $0.22 per 1M tokens → $0.44 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash Vision Exp — $0.0069 per 1M tokens → $0.01 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash Vision Exp — output price $0.65 per 1M tokens → $1.32 per 1M tokens ·
- Listed value change: Reflection: Beam 1.0 501B A23B — 131,072 tokens → 943,718 tokens ·
- Listed value change: MiniMax: MiniMax M2.7 — 131,072 tokens → 176,947 tokens ·
- Listed value change: Meta: Llama 3.3 70B Instruct — 16,384 tokens → 32,768 tokens ·
- Listed value change: Inference.net: Schematron V2 Small (inference-net) — 4,096 tokens → 8,192 tokens ·
- Listed value change: DeepSeek: DeepSeek V4.1 Flash (batch) — 384,000 tokens → 131,072 tokens ·
- Listed value change: DeepSeek: DeepSeek V4 Flash Vision Exp — 262,144 tokens → 943,717 tokens ·
- Added to tracking: inclusionAI: Ling 3.0 Tiny. First check: $0.02 per 1M tokens ·
- Added to tracking: Google: Gemini Omni 1.2 Flash. First check: $1.50 per 1M tokens ·
- Added to tracking: Google: Gemini Omni 1.1 Flash. First check: $1.50 per 1M tokens ·
- Price change: Qwen: Qwen3 235B A22B Instruct 2507 — output price $0.55 per 1M tokens → $0.80 per 1M tokens ·
- Price change: Qwen: Qwen3 235B A22B Instruct 2507 — input price $0.09 per 1M tokens → $0.14 per 1M tokens ·
- Price change: Meta: Llama 4 Maverick — $0.05 per 1M tokens → $0.17 per 1M tokens ·
- Price change: Meta: Llama 4 Maverick — output price $0.65 per 1M tokens → $1.00 per 1M tokens ·
- Price change: Meta: Llama 4 Maverick — input price $0.19 per 1M tokens → $0.35 per 1M tokens ·
- Listed value change: Xiaomi: MiMo-V2.5 — context length 1,050,000 tokens → 1,000,000 tokens ·
- Listed value change: Qwen: Qwen3 235B A22B Instruct 2507 — 16,384 tokens → 117,964 tokens ·
- Listed value change: Meta: Llama 4 Maverick — 16,384 tokens → 32,768 tokens ·
- Added to tracking: Tencent: Hy4. First check: $0.42 per 1M tokens ·
- Terms answer changed: HeyGen Terms of Service — Automation banned? Yes → No clause found ·
- Terms answer changed: HeyGen Terms of Service — Benchmarks restricted? Yes → No clause found ·
- Terms answer changed: AWS Service Terms — When was the AWS Service Terms last updated? 2026-10-30 → 2026-11-03 ·
- Price change: MoonshotAI: Kimi K3 (Moonshot AI) — output price $15.00 per 1M tokens → $16.00 per 1M tokens ·
- Price change: Mistral: Mistral Large 4 (Mistral AI) — official output price $4.18 per 1M tokens → $2.10 per 1M tokens ·
- Price change: Mistral: Mistral Large 4 (Mistral AI) — $0.14 per 1M tokens → $0.07 per 1M tokens ·
- Price change: Mistral: Mistral Large 4 (Mistral AI) — official input price $1.36 per 1M tokens → $0.70 per 1M tokens ·
- Price change: Meta: Llama 3.1 70B Instruct — output price $0.40 per 1M tokens → $0.72 per 1M tokens ·
- Price change: Meta: Llama 3.1 70B Instruct — input price $0.40 per 1M tokens → $0.72 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4.1 Flash 0910 — $0.0060 per 1M tokens → $0.02 per 1M tokens ·
- Listed value change: NVIDIA: Nemotron 3.5 Super VL — 235,929 tokens → 943,718 tokens ·
- Listed value change: Meta: Llama 3.1 70B Instruct — 16,384 tokens → 8,192 tokens ·
- Added to tracking: xAI: Grok 4.7 (batch). First check: $1.60 per 1M tokens ·
- Added to tracking: xAI: Grok 4.6 (batch). First check: $1.60 per 1M tokens ·
- Added to tracking: Crouton Alpha (stealth). First check: $0.00 per 1M tokens (free) ·
- Terms answer changed: Alibaba Cloud Product Terms of Service (Model Studio section) — When was the Alibaba Cloud Product Terms of Service (Model Studio section) last updated? 2026-09-23 → 2026-10-19 ·
- Price change: Z.ai: GLM 5.3 FlashX — $0.09 per 1M tokens → $0.08 per 1M tokens ·
- Price change: OpenAI: gpt-oss-20b — $0.0090 per 1M tokens → $0.03 per 1M tokens ·
- Price change: OpenAI: gpt-oss-20b — output price $0.09 per 1M tokens → $0.13 per 1M tokens ·
- Price change: OpenAI: gpt-oss-20b — input price $0.02 per 1M tokens → $0.03 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash 0731 — $0.02 per 1M tokens → $0.04 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash 0731 — input price $0.12 per 1M tokens → $0.06 per 1M tokens ·
- Price change: DeepSeek: DeepSeek V4 Flash 0731 — output price $0.24 per 1M tokens → $0.30 per 1M tokens ·
- Listed value change: Qwen: Qwen3.8 27B — context length 1,000,000 tokens → 1,024,000 tokens ·
- Listed value change: OpenAI: gpt-oss-20b — 32,768 tokens → 117,964 tokens ·
- Added to tracking: Z.ai: GLM 5.3 FlashX Legal. First check: $0.37 per 1M tokens ·
Archive by month
- November 2026 — 121 entries
- October 2026 — 179 entries
How changes are detected
New and removed models and tools come from our daily discovery pass over public model listings and vendor pages. The date is when we first (or last) saw them, not the vendor’s release date; they are listings, not changes. A change is listed only when the new value of the same question for the same model, tool or plan appears again in a check on a later day, with no other value in between (confirmed); a value that returns to the old one within 7 days is not a change. Each change has one ID shared with the object page, which shows the old and the new value. Price moves under 1%, values that flip back and forth between two figures, and GPU rental prices on marketplaces are not shown as changes. Limit wording is compared as text after removing spacing, quote and case differences, not by the page as a whole. Terms count as changed only when the answer itself (yes, no, no clause found, a date) changes. A failed check is never reported as a change. A plan or model added to tracking later is listed as “Added to tracking”, not as a change.
“Today” is the date of our latest check.