OpenRouter API Pricing Update: 24 Free Models, DeepSeek Price Shifts, and New Paid Entrants

As of 2026-09-24, there are 24 free models on OpenRouter, up from the previous day. These include Google: Lyria 3 Pro Preview and Google: Lyria 3 Clip Preview, both with 1,048,576 context tokens. Thinking Machines offers two free models: Inkling Small (free) and Inkling (free), each with 1,048,576 context tokens. NVIDIA provides two free models: Nemotron 3.5 Lightning (free) and Nemotron 3 Ultra (free), each with 1,000,000 context tokens. Additional free models come from Dots Studio, inclusionAI, Nex AGI, Qwen, Poolside, Google (Gemma 4 variants), Cohere, and NVIDIA (Nemotron 3 Super and Nemotron 3 Nano Omni).

Seven new models were added to OpenRouter on 2026-09-24. Among them, AionLabs: Aion 3.5 and AionLabs: Aion 3.5 Mini are paid, with input prices of $3.0 and $0.7 per million tokens and output prices of $6.0 and $1.4 per million tokens, respectively. OpenAI: gpt-oss-120b (batch) is available at $0.0296 input and $0.136 output per million tokens with open weights. Qwen: Qwen3.8 Max Prime is priced at $4.0 input and $12.0 output per million tokens. Space Bunny Alpha is now listed as free at $0.00 input and output. Upstage: Solar Mini 4 costs $0.05 input and $0.2 output per million tokens. Z.ai: GLM 5.3 Prime is priced at $2.8 input and $8.8 output per million tokens. Two models were removed: inclusionAI: Ling 3.0 Flash VL (free) and Mistral: Devstral 2 2512.

Several DeepSeek models experienced price updates on 2026-09-24. DeepSeek: DeepSeek Pro Latest decreased input price from $0.40 to $0.3865 per million tokens and output price from $4.30 to $2.90 per million tokens. DeepSeek: DeepSeek Flash Latest saw a slight input decrease from $0.10 to $0.099 per million tokens and an output increase from $0.50 to $0.60 per million tokens. DeepSeek: DeepSeek V4 Flash Latest increased input from $0.03 to $0.038 per million tokens and decreased output from $0.80 to $0.55 per million tokens. DeepSeek: DeepSeek V4 Pro 0813 reduced input from $0.66 to $0.462 per million tokens and output from $1.98 to $1.386 per million tokens. NVIDIA: Nemotron 3.5 Lightning increased input from $0.07 to $0.08 per million tokens, with output unchanged at $0.20. Z.ai: GLM 5.3 (batch) reduced input from $0.72 to $0.45 per million tokens and output from $2.40 to $2.00 per million tokens. MoonshotAI: Kimi Latest increased input from $1.35 to $1.4989 per million tokens and decreased output from $14.33 to $10.758 per million tokens. DeepSeek: DeepSeek V4.1 Flash increased input from $0.10 to $0.15 per million tokens and output from $0.50 to $0.60 per million tokens. DeepSeek: DeepSeek V4 Pro 0423 saw minor decreases: input from $0.9553 to $0.9396 per million tokens and output from $1.9105 to $1.8792 per million tokens.

The lowest input-cost paid models on OpenRouter as of 2026-09-24 are led by IBM: Granite 4.0 Micro at $0.017 per million input tokens and $0.112 output, with 131,000 context tokens and open weights. OpenAI: gpt-oss-20b follows at $0.018 input and $0.09 output per million tokens, also with open weights and 131,072 context tokens. Mistral: Mistral Nemo is priced at $0.019 input and $0.03 output per million tokens. inclusionAI: Ling 3.0 Flash costs $0.021 input and $0.063 output per million tokens with 262,144 context tokens. OpenAI: gpt-oss-20b (batch) is $0.024 input and $0.112 output per million tokens. Nex AGI: Nex-N2.5-Mini is $0.025 input and $0.10 output per million tokens. OpenAI: GPT-5 Nano (batch) is $0.025 input and $0.20 output per million tokens with 400,000 context tokens. Meta: Llama 3.2 1B Instruct is $0.027 input and $0.201 output per million tokens with 60,000 context tokens.