Infron has listed Moonshot AI’s Kimi K3 as a serverless model, while Moonshot AI describes Kimi K3 as a 2.8-trillion-parameter native-vision model with a 1M-token context window and open-weight availability.
Moonshot AI has released Kimi K3, and Infron has listed the model as available through its serverless model catalog under the ID moonshotai/kimi-k3.
Infron’s model page lists Moonshotai: Kimi K3 with the model ID moonshotai/kimi-k3, serverless availability, a 1M-token context window, and a release date of July 17, 2026. The Infron listing shows list pricing of $3.00 per million input tokens and $15.00 per million output tokens, and includes a provider table with a ZDR column.
Those prices match the public API rates Moonshot AI gives for Kimi K3’s uncached input and output in its own launch materials. Moonshot AI’s blog lists Kimi K3 pricing at $0.30 per million cached input tokens, $3.00 per million input tokens, and $15.00 per million output tokens.
Moonshot AI describes Kimi K3 as its “most capable model” and says it is a 2.8-trillion-parameter native-vision model with a 1M-token context window. The company says the model is available through Kimi.com, Kimi Work, Kimi Code, and the Kimi API.
The official Moonshot AI GitHub repository describes Kimi K3 as an open-weight native multimodal agentic model. According to that repository, Kimi K3 has 2.8 trillion total parameters, 104 billion activated parameters, and a 1,048,576-token context length. The repository also says the model uses MXFP4/MXFP8 quantization and can be accessed through Moonshot AI’s API using the model name kimi-k3.
The combination of a large context window, multimodal design, and open-weight release places Kimi K3 in the same broad category as other frontier-oriented models aimed at coding, agentic tasks, long-context reasoning, and vision-language workloads. Moonshot AI’s own materials emphasize native vision support and broad product availability rather than limiting the release to a single chat interface.
The Associated Press reported that Moonshot AI’s newest Kimi K3 model appeared to be catching up to leading models from Anthropic and OpenAI. AP also reported that Kimi K3 topped Arena’s front-end coding ranking, a notable result for developers tracking model performance on interface-building tasks.
That outside coverage does not independently establish superiority across all tasks, and benchmark rankings can change as new models and evaluations appear. Still, AP’s report indicates that Kimi K3 has drawn attention beyond Moonshot AI’s own release channels, particularly for coding-related performance.
For developers and teams already using multi-model infrastructure, Infron’s listing provides another access route for Kimi K3 alongside Moonshot AI’s own API and products. The Infron page identifies the model clearly, lists serverless availability, and publishes pricing and context-window details in a format suitable for comparing it with other hosted models.
The key verified facts are straightforward: Moonshot AI says Kimi K3 is a 2.8T-parameter native-vision model with a 1M-token context window; its official repository describes it as open-weight and multimodal; Moonshot AI lists API pricing at $0.30 cached input, $3.00 input, and $15.00 output per million tokens; and Infron now lists moonshotai/kimi-k3 with serverless availability and matching list input and output prices.
Moonshot AI has released Kimi K3, and Infron has listed the model as available through its serverless model catalog under the ID moonshotai/kimi k3 .
Kimi K3 arrives on Infron Infron’s model page lists Moonshotai: Kimi K3 with the model ID moonshotai/kimi k3 , serverless availability, a 1M token context window, and a release date of July 17, 2026.
The Infron listing shows list pricing of $3.00 per million input tokens and $15.00 per million output tokens , and includes a provider table with a ZDR column.
Continue reading
