← variables.md

2026-08-25

101 change(s) in the model market

added to the catalog

• MiniMax: MiniMax M2.7 (free) (minimax/minimax-m2.7:free) added to the catalog: 196,608 ctx, $0/M in / $0/M out
• Mistral: Devstral 2 2512 (mistralai/devstral-2512) added to the catalog: 262,144 ctx, $0.44/M in / $2.2/M out

removed from the catalog

• inclusionAI: Ling-2.6-1T (inclusionai/ling-2.6-1t) removed from the catalog
• inclusionAI: Ling-2.6-flash (inclusionai/ling-2.6-flash) removed from the catalog
• inclusionAI: Ring-2.6-1T (inclusionai/ring-2.6-1t) removed from the catalog
• NVIDIA: Nemotron 3 Nano 30B A3B (free) (nvidia/nemotron-3-nano-30b-a3b:free) removed from the catalog
• NVIDIA: Nemotron Nano 12B 2 VL (free) (nvidia/nemotron-nano-12b-v2-vl:free) removed from the catalog
• NVIDIA: Nemotron Nano 9B V2 (free) (nvidia/nemotron-nano-9b-v2:free) removed from the catalog

model price changes

• ~deepseek/deepseek-v4-flash-latest: prompt price $0.04/M → $0.035/M (−12.5%)
• ~deepseek/deepseek-v4-flash-latest: completion price $0.08/M → $0.075/M (−6.3%)
• ~moonshotai/kimi-latest: prompt price $2.6/M → $2.8/M (+7.7%)
• ~moonshotai/kimi-latest: completion price $13/M → $14/M (+7.7%)
• deepseek/deepseek-v4-flash: prompt price $0.057/M → $0.089/M (+54.4%)
• deepseek/deepseek-v4-flash: completion price $0.115/M → $0.177/M (+54.4%)
• deepseek/deepseek-v4-pro: prompt price $0.526/M → $0.79/M (+50.2%)
• deepseek/deepseek-v4-pro: completion price $1.05/M → $1.58/M (+50.2%)
• minimax/minimax-m2.7: prompt price $0.24/M → $0.3/M (+25.0%)
• minimax/minimax-m2.7: completion price $0.96/M → $1.2/M (+25.0%)
• moonshotai/kimi-k2.5: prompt price $0.45/M → $0.6/M (+33.3%)
• moonshotai/kimi-k2.5: completion price $2.25/M → $3/M (+33.3%)
• thinkingmachines/inkling: prompt price $1/M → $0.95/M (−5.0%)
• z-ai/glm-5.1: prompt price $0.966/M → $1.26/M (+30.4%)
• z-ai/glm-5.1: completion price $3.04/M → $3.96/M (+30.4%)
• z-ai/glm-5.2: prompt price $0.966/M → $1.19/M (+23.2%)
• z-ai/glm-5.2: completion price $3.04/M → $3.74/M (+23.2%)

model context changes

• qwen/qwen3-235b-a22b-thinking-2507: context length 262,144 → 131,072
• thinkingmachines/inkling-small:free: context length 262,144 → 1,048,576
• thinkingmachines/inkling:free: context length 262,144 → 1,048,576

quantization changes

• GMICloud: deepseek/deepseek-v4-flash-0731 quantization fp8 → fp4
• CoreWeave: google/gemma-4-31b-it quantization bf16 → fp4

provider price changes

• GMICloud: deepseek/deepseek-v4-flash prompt price $0.059/M → $0.112/M (+90.5%)
• GMICloud: deepseek/deepseek-v4-flash completion price $0.118/M → $0.224/M (+90.5%)
• StreamLake: deepseek/deepseek-v4-flash prompt price $0.057/M → $0.089/M (+54.4%)
• StreamLake: deepseek/deepseek-v4-flash completion price $0.115/M → $0.177/M (+54.4%)
• GMICloud: deepseek/deepseek-v4-flash-0731 prompt price $0.067/M → $0.112/M (+66.7%)
• GMICloud: deepseek/deepseek-v4-flash-0731 completion price $0.134/M → $0.224/M (+66.7%)
• Mancer 2: deepseek/deepseek-v4-flash-0731 prompt price $0.14/M → $0.15/M (+7.1%)
• OpenInference: deepseek/deepseek-v4-flash-0731 prompt price $0.04/M → $0.035/M (−12.5%)
• OpenInference: deepseek/deepseek-v4-flash-0731 completion price $0.13/M → $0.075/M (−42.3%)
• StreamLake: deepseek/deepseek-v4-flash-0731 prompt price $0.207/M → $0.088/M (−57.4%)
• StreamLake: deepseek/deepseek-v4-flash-0731 completion price $0.15/M → $0.064/M (−57.4%)
• GMICloud: deepseek/deepseek-v4-pro prompt price $0.528/M → $0.792/M (+50.0%)
• GMICloud: deepseek/deepseek-v4-pro completion price $1.58/M → $2.38/M (+50.0%)
• StreamLake: deepseek/deepseek-v4-pro prompt price $0.526/M → $0.79/M (+50.2%)
• StreamLake: deepseek/deepseek-v4-pro completion price $1.05/M → $1.58/M (+50.2%)
• Alibaba: deepseek/deepseek-v4-pro-0813 prompt price $1.16/M → $1.12/M (−3.4%)
• Alibaba: deepseek/deepseek-v4-pro-0813 completion price $3.48/M → $3.37/M (−3.4%)
• GMICloud: minimax/minimax-m2.7 prompt price $0.24/M → $0.3/M (+25.0%)
• GMICloud: minimax/minimax-m2.7 completion price $0.96/M → $1.2/M (+25.0%)
• GMICloud: minimax/minimax-m3 prompt price $0.24/M → $0.6/M (+150.0%)
• GMICloud: minimax/minimax-m3 completion price $0.96/M → $2.4/M (+150.0%)
• Mistral: mistralai/mistral-small-2603 prompt price $0.165/M → $0.15/M (−9.1%)
• Mistral: mistralai/mistral-small-2603 completion price $0.66/M → $0.6/M (−9.1%)
• Inceptron: moonshotai/kimi-k2.6 prompt price $0.6/M → $0.57/M (−5.0%)
• Alibaba: qwen/qwen3.8-27b prompt price $0.5/M → $0.425/M (−15.0%)
• Alibaba: qwen/qwen3.8-27b completion price $3/M → $2.55/M (−15.0%)
• Mancer 2: z-ai/glm-4.7 prompt price $0.6/M → $0.65/M (+8.3%)
• GMICloud: z-ai/glm-5.2 prompt price $1.4/M → $1.05/M (−25.0%)
• GMICloud: z-ai/glm-5.2 completion price $4.4/M → $3.3/M (−25.0%)

provider context changes

• GMICloud: deepseek/deepseek-v4-flash-0731 context length 1,048,575 → 1,048,576

providers added

• Mistral now serves mistralai/codestral-2508 (unknown, 256,000 ctx, $0.33/M in / $0.99/M out)
• Mistral now serves mistralai/ministral-14b-2512 (unknown, 262,144 ctx, $0.2/M in / $0.2/M out)
• Mistral now serves mistralai/ministral-14b-2512 (unknown, 262,144 ctx, $0.22/M in / $0.22/M out)
• Mistral now serves mistralai/ministral-3b-2512 (unknown, 131,072 ctx, $0.1/M in / $0.1/M out)
• Mistral now serves mistralai/ministral-3b-2512 (unknown, 131,072 ctx, $0.11/M in / $0.11/M out)
• Mistral now serves mistralai/ministral-8b (unknown, 128,000 ctx, $0.11/M in / $0.11/M out)
• Mistral now serves mistralai/ministral-8b-2512 (unknown, 262,144 ctx, $0.15/M in / $0.15/M out)
• Mistral now serves mistralai/ministral-8b-2512 (unknown, 262,144 ctx, $0.165/M in / $0.165/M out)
• Mistral now serves mistralai/mistral-large-2512 (unknown, 262,144 ctx, $0.5/M in / $1.5/M out)
• Mistral now serves mistralai/mistral-large-2512 (unknown, 262,144 ctx, $0.55/M in / $1.65/M out)
• Mistral now serves mistralai/mistral-medium-3 (unknown, 131,072 ctx, $0.4/M in / $2/M out)
• Mistral now serves mistralai/mistral-medium-3 (unknown, 131,072 ctx, $0.44/M in / $2.2/M out)
• Mistral now serves mistralai/mistral-medium-3-5 (unknown, 262,144 ctx, $1.5/M in / $7.5/M out)
• Mistral now serves mistralai/mistral-medium-3-5 (unknown, 262,144 ctx, $1.65/M in / $8.25/M out)
• Mistral now serves mistralai/mistral-medium-3.1 (unknown, 131,072 ctx, $0.4/M in / $2/M out)
• Mistral now serves mistralai/mistral-medium-3.1 (unknown, 131,072 ctx, $0.44/M in / $2.2/M out)
• Mistral now serves mistralai/mistral-nemo (unknown, 131,072 ctx, $0.165/M in / $0.165/M out)
• Mistral now serves mistralai/mistral-small-2603 (unknown, 262,144 ctx, $0.165/M in / $0.66/M out)
• Mistral now serves mistralai/mistral-small-2603 (unknown, 262,144 ctx, $0.165/M in / $0.66/M out)
• Mistral now serves mistralai/mistral-small-3.2-24b-instruct (unknown, 131,072 ctx, $0.11/M in / $0.33/M out)
• GMICloud now serves moonshotai/kimi-k2.6 (fp8, 262,144 ctx, $0.855/M in / $3.6/M out)
• Fireworks now serves moonshotai/kimi-k3 (unknown, 1,048,576 ctx, $4.5/M in / $22.5/M out)
• Nebius now serves nvidia/nemotron-3-nano-30b-a3b (fp8, 262,144 ctx, $0.06/M in / $0.24/M out)
• OpenAI now serves openai/gpt-5.2 (unknown, 400,000 ctx, $3.5/M in / $28/M out)
• OpenAI now serves openai/gpt-5.3-codex (unknown, 400,000 ctx, $3.5/M in / $28/M out)
• OpenAI now serves openai/gpt-5.5 (unknown, 1,050,000 ctx, $12.5/M in / $75/M out)
• Cloudflare now serves qwen/qwen3.8-27b (unknown, 262,144 ctx, $0.45/M in / $3.2/M out)
• Phala now serves qwen/qwen3.8-27b (unknown, 262,144 ctx, $0.4/M in / $3/M out)
• Io Net now serves z-ai/glm-5.2 (fp8, 262,144 ctx, $1.26/M in / $4.4/M out)
• Mistral now serves z-ai/glm-5.2 (unknown, 1,048,576 ctx, $1.4/M in / $4.4/M out)
• Mistral now serves z-ai/glm-5.2 (unknown, 1,048,576 ctx, $1.54/M in / $4.84/M out)

providers removed

• Decart no longer serves deepseek/deepseek-v4-flash-0731 (was fp4)
• Sail Research no longer serves deepseek/deepseek-v4-flash-0731 (was fp4)
• Mistral no longer serves mistralai/ministral-8b (was unknown)
• Sail Research no longer serves moonshotai/kimi-k3 (was fp4)
• Wafer no longer serves moonshotai/kimi-k3 (was unknown)
• OpenAI no longer serves openai/gpt-5.2 (was unknown)
• OpenAI no longer serves openai/gpt-5.3-codex (was unknown)
• OpenAI no longer serves openai/gpt-5.5 (was unknown)
• Together no longer serves qwen/qwen-2.5-7b-instruct (was unknown)
• DeepInfra no longer serves qwen/qwen3-235b-a22b-thinking-2507 (was fp8)

openrouter catalog snapshots 2026-08-24 → 2026-08-25 · schema 1 · generated 2026-08-25T05:47:50.427Z