OpenRouter API Pricing Update: October 6, 2026

Two models were removed from the free tier: Qwen: Qwen3.8 27B (free) from provider qwen and Space Bunny Alpha from provider stealth. No models became free or stopped being free today.

Output pricing increased for Qwen: Qwen3.6 27B from $3.2 to $3.25 per million tokens. DeepSeek: DeepSeek V3 0324 input rose from $0.25 to $0.29 and output from $1.0 to $1.14 per million tokens. DeepSeek: DeepSeek Pro Latest input increased from $0.1901 to $0.2783 per million tokens with output unchanged at $4.2. DeepSeek: DeepSeek V4 Flash 0423 input rose from $0.0224 to $0.03 per million tokens, output unchanged at $1.28. Z.ai: GLM 5.3 input dropped sharply from $1.4 to $0.07 per million tokens while output increased from $4.4 to $7.0. DeepSeek: DeepSeek V3.2 input decreased from $0.28 to $0.259 per million tokens, output unchanged at $0.42. OpenAI: GPT-5.6 Sol Pro input halved from $4.0 to $2.0 and output halved from $20.0 to $10.0 per million tokens. MoonshotAI: Kimi Latest input increased from $0.6566 to $0.7959 per million tokens, output unchanged at $13.0. MoonshotAI: Kimi K3 input increased from $0.72 to $1.19 per million tokens, output unchanged at $14.0. DeepSeek: DeepSeek V4.1 Flash input surged from $0.003 to $0.3 per million tokens while output fell from $2.4 to $1.2. DeepSeek: DeepSeek V3.1 Terminus input rose from $0.27 to $0.3 per million tokens, output unchanged at $1.0. Google: Gemma 4 26B A4B input increased from $0.0675 to $0.0765 and output from $0.225 to $0.255 per million tokens.

The lowest input price among paid models is $0.003 per million tokens for DeepSeek: DeepSeek Flash Latest (~deepseek) with output at $2.4 per million tokens and 1,048,576 context tokens. Next is DeepSeek: DeepSeek V4 Flash Latest (~deepseek) at $0.0152 input and $1.28 output per million tokens with 1,048,576 context tokens. DeepSeek: DeepSeek V4 Flash 0731 (deepseek) matches at $0.0152 input and $1.28 output with open weights and same context. IBM: Granite 4.0 Micro offers $0.017 input and $0.112 output per million tokens with 131,000 context tokens and open weights. OpenAI: gpt-oss-20b provides $0.018 input and $0.09 output per million tokens with 131,072 context tokens and open weights. Mistral: Mistral Nemo has $0.019 input and $0.03 output per million tokens with 131,072 context tokens and open weights. inclusionAI: Ling 3.0 Flash VL and Ling 3.0 Flash both list at $0.021 input with $0.0616 and $0.063 output per million tokens respectively, each with 262,144 context tokens and open weights.

The largest context windows available are 2,000,000 tokens, held by Auto Router (Beta), Pareto Code Router, SpaceXAI: Grok 4.20 Multi-Agent, SpaceXAI: Grok 4.20, and Auto Router — all from providers openrouter or x-ai.