> This page is for Changelog.

> Sarvam Voice Agents (Samvaad) is a voice AI agent platform for Indian languages: build, test and deploy AI calling agents on your own phone numbers, from ₹3 a minute. Start here: https://docs.sarvam.ai/voice-agents/why-voice-agents.
> For clean Markdown of any page, append `.md` to the page URL.
> For a complete documentation index, see https://docs.sarvam.ai/llms.txt. For a section index, append `/llms.txt` to any section URL (for example https://docs.sarvam.ai/voice-agents/llms.txt).
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.

## New Models on Chat Completion V2 and Responses

Model APIs

New open-weight models are available on `POST /v2/chat/completions` and `POST /v2/responses`.

**Generally available:**

* **[`glm5.3-flash`](/api/token-factory/models/glm-5-3-flash)**: a faster GLM-5.3 with image input, log probabilities, and `n` > 1. Reasons by default. 1,048,576-token context.

**In beta:**

* **[`deepseekv4.1-flash-1017`](/api/token-factory/models/deepseek-v4.1-flash-1017)**: DeepSeek V4.1 Flash 1017, low-cost reasoning with image input for long documents and agents. Reasons at `high` by default; `reasoning_effort: "none"` turns thinking off. 1,048,576-token context.
* **[`deepseekv4-pro`](/api/token-factory/models/deepseek-v4-pro)**: the larger DeepSeek V4 model for text, with tool calling and structured output. Reasons at `high` by default; `reasoning_effort: "none"` turns thinking off. 1,000,000-token context.
* **[`kimi-k3`](/api/token-factory/models/kimi-k3)**: always-on reasoning with image input and short video clips (beta), sent as base64. 1,048,576-token context.
* **[`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash)**: the lowest-cost open-weight model, with image, audio, and video input (base64; video on chat only), tool calling, and structured output. Reasons by default; `reasoning_effort: "none"` turns thinking off. 1,048,576-token context.
* **[`mimo-v2.6-pro`](/api/token-factory/models/mimo-v2.6-pro)**: the larger MiMo V2.6 model, with the same API behaviour as MiMo V2.6 Flash. 1,048,576-token context.

All models take images, audio, and video as inline base64 data URIs; remote media URLs return `400`. Requests can be up to 40 MiB on `kimi-k3` and the MiMo models, and 10 MiB on other models.

See [Beta models](/api-reference/beta-apis).

## Chat Completion V2 and Responses Are Generally Available

Model APIs

`POST /v2/chat/completions` and `POST /v2/responses` are now generally available with `sarvam-105b`, `glm5.3`, `glm5.3-flash`, and `gemma4`. Your API key no longer needs to be whitelisted. The new models in the entry above are in beta; see [Beta models](/api-reference/beta-apis).