September 28, 2026
Meta brings AI tools for businesses into one platform, led by former MongoDB CEO
On Sep 28, Meta launched Meta Enterprise Platform: its first phase includes the Muse agent, Meta Business Agent, Muse API, and Muse Code. The division will be led by Chirantan Desai, known as CJ, who previously led MongoDB. He will report directly to Mark Zuckerberg.
Muse Code entered beta on Aug 5. On Sep 24, Meta took it out of beta and added Windows, and has now placed it alongside an agent, API, and business tools. There is no unified package available to buy yet: Meta has not described its price, availability, how the products work together, or enterprise access controls.
Muse Code can already be installed on macOS and Linux with `curl -fsSL https://dev.meta.ai/install.sh | sh`. In the project directory, run `muse` and sign in through a browser or with an API key.
Code and API are already available. Meta Model API works with the OpenAI SDK, Anthropic SDK, and OpenAI-compatible agent CLI. Direct calls require a key in `MODEL_API_KEY` and the base URL `https://api.meta.ai/v1`. Meta's example uses `muse-spark-1.3` with a context window of 1,048,576 tokens and responses of up to 131,072 tokens.
Meta Model API pricing. - Standard: $1.25 per 1 million input tokens, $4.25 per 1 million output tokens, and $0.15 for cached input. Meta does not use prompts and completions from this tier to train models. - Contributor: $0.10 per 1 million input tokens and $0.20 per 1 million output tokens. Meta may use prompts and completions to train future models.
Under Standard, the team gets a limit of 3,000 RPM and 4,000,000 TPM. Under Contributor, the limit is 100 RPM and 3,000,000 TPM, with all of the team's API keys sharing one quota.
Chirantan Desai will report directly to Mark Zuckerberg.