DeepSeek V4.1 Flash input price jumps to $0.30 per million tokens
Por ahora este informe solo está disponible en inglés.
The most significant pricing shift on OpenRouter today is the increase in DeepSeek V4.1 Flash input costs from $0.051 to $0.30 per million tokens. This change affects one of the most cost-effective high-context options available. The platform now tracks 475 models, with 18 offering free access and 192 featuring open weights. No models were removed from the platform, and no previously paid models became free. The free tier remains stable at 18 models, including new additions from Particle AI and Thinking Machines.
Major Price Increases
DeepSeek raised the input price for DeepSeek V4.1 Flash from $0.051 to $0.30 per million tokens, while output costs rose from $0.567 to $1.20 per million tokens. The batch version of the same model saw its input price increase from $0.011 to $0.05 per million tokens, with output costs dropping from $0.20 to $0.15 per million tokens.
Google: Gemini Flash Latest doubled its input price from $0.75 to $1.50 per million tokens, and output costs increased from $3.75 to $7.50 per million tokens. OpenAI: GPT-5.6 Sol saw a modest increase, with input prices rising from $4.00 to $4.40 per million tokens and output prices moving from $20.00 to $22.00 per million tokens.
Price Decreases and Adjustments
Qwen: Qwen3.5-9B reduced its input price from $0.17 to $0.082 per million tokens, and output costs fell from $0.25 to $0.12 per million tokens. IBM: Granite 4.2 8B kept its input price at $0.06 per million tokens but lowered output costs from $0.25 to $0.15 per million tokens.
MoonshotAI: Kimi K2.5 decreased its input price from $0.532 to $0.50 per million tokens and cut output costs from $3.325 to $2.50 per million tokens. Z.ai: GLM 5.2 (batch) lowered its input price from $0.527 to $0.322 per million tokens and output costs from $1.075 to $0.989 per million tokens.
New Model Additions
Eight new models joined the platform today. Google: Gemini 4 Hydrogen launched with an input price of $1.50 and an output price of $7.50 per million tokens. Its batch variant is priced at $0.75 for input and $3.75 for output per million tokens.
inclusionAI added two models, Ling 3.0 Flash Fin and Ling 3.0 Flash Sante, both priced at $0.042 for input and $0.1232 for output per million tokens. Particle AI: Particle-1 (free) and Thinking Machines: Inkling 2 Small (free) were added to the free tier, with the latter also available as a paid option at $0.30 for input and $1.00 for output per million tokens.
Cheapest Paid Options
The lowest input price among paid models is $0.007 per million tokens for Z.ai: GLM 5.2, though its output price is $6.80 per million tokens. inclusionAI: Ling 3.0 Tiny follows at $0.015 for input and $0.045 for output per million tokens.
Meta: Llama 3.1 8B Instruct is priced at $0.02 for input and $0.04 for output per million tokens. DeepSeek: DeepSeek Flash Latest costs $0.022 for input and $0.40 for output per million tokens. OpenAI: GPT-5 Nano (batch) is available for $0.025 for input and $0.20 for output per million tokens.
Puntos clave
- DeepSeek V4.1 Flash input costs increased by a factor of nearly six, moving from $0.051 to $0.30 per million tokens.
- Google: Gemini Flash Latest doubled its input and output prices, reaching $1.50 and $7.50 per million tokens respectively.
- The free tier remains constant at 18 models, with no models transitioning between free and paid status today.
- Z.ai: GLM 5.2 holds the lowest input price at $0.007 per million tokens, despite a high output cost of $6.80.
- New additions include Google: Gemini 4 Hydrogen and two inclusionAI models, expanding the mid-range pricing options.