OpenRouter API Pricing Update: October 2, 2026

As of October 2, 2026, there are 21 free models available on OpenRouter, unchanged from the previous day. The free models list includes Google's Lyria 3 Pro Preview and Lyria 3 Clip Preview, both with 1,048,576 context tokens, and Thinking Machines' Inkling Small and Inkling models, also free with 1,048,576 context tokens. Additional free models come from NVIDIA, Poolside, inclusionAI, Qwen, Cohere, and others, with context lengths ranging from 128,000 to 1,048,576 tokens. The Free Models Router from OpenRouter remains available with 200,000 context tokens.

Two new models were added on October 2, 2026: Apodex: Apodex 1.1 Mini (free) from provider apodex, with zero input and output pricing and 262,144 context tokens, and Pareto 26.10 Preview from provider unbiased, priced at $0.80 per million input tokens and $3.20 per million output tokens with 1,048,576 context tokens. Two models were removed: OpenAI's GPT-6.1 Sol Pro (batch) and GPT-6.1 Sol (batch), both from the openai provider. No models became free or stopped being free on this day.

Several models experienced price changes on October 2, 2026. DeepSeek: DeepSeek Pro Latest increased input pricing from $0.123 to $0.1431 per million tokens while output remained at $3.50. DeepSeek: DeepSeek Flash Latest raised input from $0.0198 to $0.0264 and output from $0.396 to $0.60 per million tokens. Conversely, DeepSeek: DeepSeek V4 Flash 0423 reduced input from $0.0785 to $0.0419 and output from $0.1571 to $0.0837 per million tokens. Z.ai: GLM 5.3 saw a significant drop in input from $1.40 to $0.2219 and output from $4.40 to $3.39 per million tokens. MoonshotAI: Kimi Latest and Kimi K3 both increased input from $0.28 to $0.4357 per million tokens with output unchanged at $10.00, while MoonshotAI: Kimi K2.6 reduced input from $0.65 to $0.4341 and output from $3.41 to $1.828 per million tokens. MiniMax: MiniMax M1 increased input from $0.40 to $0.55 per million tokens with output steady at $2.20.

The lowest input-priced paid models on October 2, 2026 are DeepSeek: DeepSeek V4 Flash Latest and DeepSeek: DeepSeek V4 Flash 0731, both at $0.0108 per million input tokens and $1.28 per million output tokens with 1,048,576 context tokens. The DeepSeek V4 Flash 0731 variant is noted as open weights, while the Latest version is not. Next in affordability are IBM: Granite 4.0 Micro at $0.017 input and $0.112 output, OpenAI: gpt-oss-20b at $0.018 input and $0.09 output, and Mistral: Mistral Nemo at $0.019 input and $0.03 output, all with 131,000–131,072 context tokens and open weights. Z.ai: GLM Flash Latest offers $0.02 input and $0.30 output with 1,048,576 context tokens but is not open weights.