OpenRouter API Pricing Update: October 10, 2026

There are 19 free models available as of 2026-10-10, including Google: Lyria 3 Pro Preview and Google: Lyria 3 Clip Preview, each with 1,048,576 context tokens. Thinking Machines: Inkling Small (free) and Thinking Machines: Inkling (free) also offer 1,048,576 context tokens. NVIDIA: Nemotron 3.5 Lightning (free) and NVIDIA: Nemotron 3 Ultra (free) provide 1,000,000 context tokens.

No new models were added on 2026-10-10. Ten Qwen models were removed, including Qwen: Qwen Plus 0728, Qwen: Qwen3 235B A22B, Qwen: Qwen3 30B A3B Thinking 2507, Qwen: Qwen3 8B, Qwen: Qwen3 Coder Plus, Qwen: Qwen3 Max, Qwen: Qwen3 Max Thinking, Qwen: Qwen3 VL 235B A22B Thinking, Qwen: Qwen3 VL 32B Instruct, and Qwen: Qwen3 VL 8B Thinking.

Several models experienced price changes on 2026-10-10. Qwen: Qwen3 Next 80B A3B Instruct increased input price from $0.09 to $0.10 per million tokens, with output unchanged at $1.10. Qwen: Qwen3.6 27B raised input from $0.30 to $0.32 and output from $2.00 to $3.20 per million tokens. Qwen: Qwen3.5-35B-A3B reduced input from $0.15 to $0.08 and output from $1.00 to $0.75 per million tokens. Z.ai: GLM 4.6 increased input from $0.43 to $0.50 and output from $1.75 to $2.00 per million tokens.

The lowest input-priced paid model is DeepSeek: DeepSeek V4 Flash Latest at $0.0144 per million tokens, with output at $1.536 per million tokens and 1,048,576 context tokens. DeepSeek: Flash Latest follows at $0.016 input and $0.60 output per million tokens with the same context length. IBM: Granite 4.0 Micro offers $0.017 input and $0.112 output per million tokens with 131,000 context tokens and open weights.