LLM API Pricing Report — 2026-11-01

Seventeen models are listed as free on OpenRouter. Google provides Lyria 3 Pro Preview and Lyria 3 Clip Preview, each with a 1,048,576-token context window. Thinking Machines offers Inkling Small and Inkling, both with 1,048,576 tokens. NVIDIA contributes five free models: Nemotron 3.5 Super VL, Nemotron 3.5 Lightning, Nemotron 3 Ultra (each 1,000,000 tokens), Nemotron 3 Nano Omni (256,000 tokens), and Nemotron 3.5 Content Safety (128,000 tokens). Dots Studio provides Dots3-Note Preview at 512,000 tokens. Recursive offers Flint at 262,144 tokens. XingChen provides Xing4.0 29B A4B at 262,144 tokens. Google adds Gemma 4 26B A4B and Gemma 4 31B, each at 262,144 tokens. Cohere offers North Mini Code at 256,000 tokens. The Free Models Router from OpenRouter supports 200,000 tokens. LiquidAI provides LFM2.5-2.6B at 65,536 tokens.

Four Poolside models were removed from the catalog: Laguna S 2.1, Laguna S 2.1 (free), Laguna XS 2.1, and Laguna XS 2.1 (free). No models were added, became free, or stopped being free since the previous day.

Twelve models had price adjustments. Z.ai GLM 5.3 input fell from $0.08 to $0.0699 per million tokens; output dropped from $3.184 to $1.40. Z.ai GLM 4.7 input rose from $0.54 to $0.60; output rose from $1.98 to $2.20. DeepSeek Pro Latest output doubled from $4.00 to $8.00 per million tokens while input held at $0.1035. Z.ai GLM 5.2 output decreased from $7.00 to $6.80 with input unchanged at $0.006. DeepSeek V4 Flash Latest input fell from $0.056 to $0.0434; output unchanged at $0.30. Qwen 3.6 27B input rose from $0.32 to $0.45; output fell from $3.20 to $2.70. DeepSeek Flash Latest input dropped from $0.022 to $0.0118; output fell from $0.192 to $0.091. MiniMax M2.5 input decreased from $0.30 to $0.27; output from $1.20 to $0.95. Qwen 3.8 27B input fell sharply from $0.15 to $0.04; output rose from $1.875 to $2.30. DeepSeek V4 Pro 0813 input dropped from $0.66 to $0.115; output rose from $1.98 to $4.00. MoonshotAI Kimi Latest input fell from $0.1615 to $0.1495; output dropped from $10.5572 to $7.15. Z.ai GLM 5.3 Flash input decreased from $0.0216 to $0.0149; output rose from $0.1238 to $0.1543.

The eight lowest input-price paid models are led by Z.ai GLM 5.2 at $0.006 per million input tokens and $6.80 output, with a 1,048,576-token context and open weights. DeepSeek Flash Latest follows at $0.0118 input and $0.091 output, same context, closed weights. Z.ai GLM Flash Latest and Z.ai GLM 5.3 Flash both sit at $0.0149 input and $0.1543 output, 1,048,576 tokens; the former is closed-weight, the latter open-weight. OpenAI gpt-oss-20b is $0.018 input and $0.09 output with 131,072 tokens and open weights. InclusionAI Ling 3.0 Flash is $0.021 input and $0.063 output at 262,144 tokens, open weights. OpenAI gpt-oss-20b (batch) is $0.024 input and $0.112 output, 131,072 tokens, open weights. Nex AGI Nex-N2.5-Mini is $0.025 input and $0.10 output at 262,144 tokens, open weights.

Five models or routers offer a 2,000,000-token context window: SpaceXAI Grok 4.20 Multi-Agent, SpaceXAI Grok 4.20, Auto Router (Beta), Pareto Code Router, and Auto Router, all via OpenRouter or x-ai.