OpenRouter API Pricing Update: 24 Free Models, Price Increases for DeepSeek V3 and Others

There are 24 free models available on OpenRouter as of 2026-09-18, including Google's Lyria 3 Pro Preview and Lyria 3 Clip Preview, both with 1,048,576 context tokens.

Thinking Machines offers two free models: Inkling Small and Inkling, each with 1,048,576 context tokens. NVIDIA provides two free models: Nemotron 3.5 Lightning and Nemotron 3 Ultra, each with 1,000,000 context tokens.

Additional free models come from Dots Studio, inclusionAI (three variants of Ling 3.0 Flash), Nex AGI (two variants of Nex-N2.5), Qwen (Qwen3.8 27B), Poolside (two Laguna variants), Google (two Gemma 4 variants), NVIDIA (Nemotron 3 Super), Cohere (North Mini Code), and NVIDIA (Nemotron 3 Nano Omni).

Two new models were added: Qwen: Qwen3.8 27B (free) from provider qwen, and Pareto from provider unbiased, priced at $2.50 per million input tokens and $7.50 per million output tokens, both with 262,144 context tokens.

One model was removed: Union Alpha from provider stealth. No models became free or stopped being free compared to the previous day.

Eight models had price changes. DeepSeek: DeepSeek V3 (provider deepseek) increased input price from $0.2574 to $0.32 per million tokens and decreased output price from $1.0287 to $0.89 per million tokens.

NVIDIA: Nemotron 3 Nano 30B A3B (provider nvidia) increased input from $0.05 to $0.06 and output from $0.20 to $0.24 per million tokens. Z.ai: GLM 5.3 Flash (provider z-ai) increased input from $0.07 to $0.09 and output from $0.2333 to $0.30 per million tokens.

Z.ai: GLM Flash Latest (provider ~z-ai) increased input from $0.07 to $0.075 and output from $0.2333 to $0.25 per million tokens. NVIDIA: Nemotron 3 Ultra (provider nvidia) increased input from $0.60 to $0.625 and output from $2.40 to $3.125 per million tokens.

Meta: Muse Glimmer 30B (provider meta) decreased input from $0.35 to $0.30 and output from $1.50 to $1.10 per million tokens. Z.ai: GLM Latest (provider ~z-ai) increased input from $0.833 to $0.8775 and output from $2.618 to $2.97 per million tokens.

MoonshotAI: Kimi Latest (provider ~moonshotai) increased input from $2.00 to $2.10 and decreased output from $11.20 to $10.95 per million tokens.

The cheapest paid model by input price is IBM: Granite 4.0 Micro (provider ibm-granite) at $0.017 per million input tokens and $0.112 per million output tokens, with 131,000 context tokens and open weights.

Second is Mistral: Mistral Nemo (provider mistralai) at $0.019 per million input and $0.03 per million output, with 131,072 context tokens and open weights.

Third is inclusionAI: Ling 3.0 Flash (provider inclusionai) at $0.021 per million input and $0.063 per million output, with 262,144 context tokens and open weights.

The largest context windows available are 2,000,000 tokens, held by five models: Auto Router (Beta), Pareto Code Router, and two variants of SpaceXAI: Grok 4.20 Multi-Agent and Grok 4.20 (all from provider x-ai), and Auto Router (provider openrouter). All are provided by openrouter or x-ai.