openai
OpenAI raises prices for GPT-5.6 Sol and gpt-oss-20b on OpenRouter
Por ahora esta nota solo está disponible en inglés.
OpenAI increased the API pricing for two of its models listed on OpenRouter on November 16, 2026. The input and output costs for GPT-5.6 Sol rose by 10.0%, while the open-weight gpt-oss-20b model saw a 150.0% increase in input costs and a 130.8% increase in output costs compared to the previous day.
Precio en OpenRouter, por 1M de tokens
- 4,40 US$
- 22,00 US$
Cambios de precio en los últimos 90 días
- entrada 4,00 US$ → 4,40 US$, salida 20,00 US$ → 22,00 US$
- entrada 4,40 US$ → 4,00 US$, salida 22,00 US$ → 20,00 US$
- entrada 4,00 US$ → 4,40 US$, salida 20,00 US$ → 22,00 US$
- entrada 2,00 US$ → 4,00 US$, salida 10,00 US$ → 20,00 US$
- 0,075 US$
- 0,30 US$
Cambios de precio en los últimos 90 días
- entrada 0,03 US$ → 0,075 US$, salida 0,13 US$ → 0,30 US$
- entrada 0,018 US$ → 0,03 US$, salida 0,09 US$ → 0,13 US$
- entrada 0,03 US$ → 0,018 US$, salida 0,13 US$ → 0,09 US$
GPT-5.6 Sol Price Adjustment
The price for OpenAI: GPT-5.6 Sol, a closed-weight model with a 1,050,000 token context window, increased from $4.00 to $4.40 per million input tokens. The output price moved from $20.00 to $22.00 per million tokens.
This model was released on July 9, 2026. Among other OpenAI models currently listed on OpenRouter, GPT-5.6 Sol Pro is priced at $4.00 per million input tokens and $20.00 per million output tokens. GPT-5.5 and GPT Chat Latest are both priced at $5.00 per million input tokens, with output prices of $30.00 per million tokens.
gpt-oss-20b Cost Increase
The open-weight model OpenAI: gpt-oss-20b, which has 20.9 billion parameters and a 131,072 token context window, experienced a significant price hike. The input price rose from $0.03 to $0.075 per million tokens, and the output price increased from $0.13 to $0.30 per million tokens.
According to OpenAI, gpt-oss-20b is intended for lower latency and local or specialized use cases. The model has 21B parameters with 3.6B active parameters and is released under a permissive Apache 2.0 license. OpenAI states that MXFP4 quantization allows the model to run within 16GB of memory, and it can be fine-tuned on consumer hardware.
Puntos clave
- GPT-5.6 Sol input and output prices both increased by 10.0% on OpenRouter.
- gpt-oss-20b input prices rose by 150.0% and output prices by 130.8% on OpenRouter.
- OpenAI describes gpt-oss-20b as suitable for local use cases with 3.6B active parameters.