> This page is for Model APIs. > 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. # Migrating from Fireworks AI: Overview > Base URL, auth, model ID mapping, and SDK changes for teams moving open-weight model traffic from Fireworks AI to Sarvam's OpenAI-compatible /v2 API. *Last updated: October 23, 2026* Fireworks AI and Sarvam's `/v2` API both speak the OpenAI Chat Completions format, so most migrations are a base URL, key, and model ID change. This page covers those setup changes. The [Chat Completion](/api/token-factory/migrations/from-fireworks/chat-completion) and [Responses API](/api/token-factory/migrations/from-fireworks/responses-api) guides cover request fields, reasoning, tools, and the Fireworks-only features to remove. ## Quick start Change the client and the model ID. Use the `openai` package you already have, or Sarvam's own SDK ([SDKs & Libraries](/api/getting-started/sdks)). If you use the Fireworks SDK, the **Before (Fireworks SDK)** option shows that version; the request shape is the same, so you can move to either the `openai` package or Sarvam's SDK. #### Python #### Before (Fireworks) ```python from openai import OpenAI client = OpenAI( api_key="YOUR_FIREWORKS_API_KEY", base_url="https://api.fireworks.ai/inference/v1", ) response = client.chat.completions.create( model="accounts/fireworks/models/kimi-k3", messages=[{"role": "user", "content": "Hello!"}], ) ``` #### Before (Fireworks SDK) ```python from fireworks import Fireworks client = Fireworks(api_key="YOUR_FIREWORKS_API_KEY") response = client.chat.completions.create( model="accounts/fireworks/models/kimi-k3", messages=[{"role": "user", "content": "Hello!"}], ) ``` #### After (OpenAI SDK) ```python from openai import OpenAI client = OpenAI( api_key="YOUR_SARVAM_API_KEY", base_url="https://api.sarvam.ai/v2", ) response = client.chat.completions.create( model="kimi-k3", messages=[{"role": "user", "content": "Hello!"}], ) ``` #### After (Sarvam SDK) ```python from sarvamai import SarvamAI client = SarvamAI(api_subscription_key="YOUR_SARVAM_API_KEY") response = client.chat.completions_v2( model="kimi-k3", messages=[{"role": "user", "content": "Hello!"}], ) ``` #### JavaScript #### Before (Fireworks) ```javascript import OpenAI from "openai"; const client = new OpenAI({ apiKey: "YOUR_FIREWORKS_API_KEY", baseURL: "https://api.fireworks.ai/inference/v1", }); const response = await client.chat.completions.create({ model: "accounts/fireworks/models/kimi-k3", messages: [{ role: "user", content: "Hello!" }], }); ``` #### After (OpenAI SDK) ```javascript import OpenAI from "openai"; const client = new OpenAI({ apiKey: "YOUR_SARVAM_API_KEY", baseURL: "https://api.sarvam.ai/v2", }); const response = await client.chat.completions.create({ model: "kimi-k3", messages: [{ role: "user", content: "Hello!" }], }); ``` #### After (Sarvam SDK) ```javascript import { SarvamAIClient } from "sarvamai"; const client = new SarvamAIClient({ apiSubscriptionKey: "YOUR_SARVAM_API_KEY" }); const response = await client.chat.completionsV2({ model: "kimi-k3", messages: [{ role: "user", content: "Hello!" }], }); ``` For requests made directly over HTTP, change the URL, the auth header, and the model ID: #### Before (Fireworks) ```bash curl https://api.fireworks.ai/inference/v1/chat/completions \ -H "Authorization: Bearer YOUR_FIREWORKS_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "accounts/fireworks/models/glm-5p3-flash", "messages": [{"role": "user", "content": "Explain UPI in two sentences."}], "max_tokens": 1000 }' ``` #### After (Sarvam) ```bash curl https://api.sarvam.ai/v2/chat/completions \ -H "api-subscription-key: YOUR_SARVAM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "glm5.3-flash", "messages": [{"role": "user", "content": "Explain UPI in two sentences."}], "max_tokens": 1000 }' ``` > **Note** > > Sarvam also accepts `Authorization: Bearer `, which is what the `openai` package sends, > so you don't need to change how your client authenticates. Create a key in the > [Sarvam dashboard](https://indus.sarvam.ai/key-management?ref=docs). ## Zero-code switch The `openai` package for Python and JavaScript reads its endpoint and key from the `OPENAI_BASE_URL` and `OPENAI_API_KEY` environment variables when you create the client without them. If your app does that, and reads the model ID from configuration, you can switch without changing code: ```bash export OPENAI_BASE_URL="https://api.sarvam.ai/v2" export OPENAI_API_KEY="YOUR_SARVAM_API_KEY" ``` ```python from openai import OpenAI client = OpenAI() # picks up OPENAI_BASE_URL and OPENAI_API_KEY ``` Then set your configured model to a [Sarvam model ID](#model-ids). If your requests send Fireworks-only fields, remove them as described in the [Chat Completion guide](/api/token-factory/migrations/from-fireworks/chat-completion). This works with the `openai` package only. If your code uses the Fireworks SDK, switch to the `openai` package first, as in the [quick start](#quick-start). ## Base URL and auth | | Fireworks AI | Sarvam | | ------------------- | --------------------------------------------------------------------------------------- | -------------------------------------------------------------- | | Base URL | `https://api.fireworks.ai/inference/v1` (or `https://us.api.fireworks.ai/inference/v1`) | `https://api.sarvam.ai/v2` | | Chat Completions | `POST /chat/completions` | `POST /chat/completions` | | Responses | `POST /responses` | `POST /responses` (stateless) | | Auth header | `Authorization: Bearer ` | `api-subscription-key: ` or `Authorization: Bearer ` | | Auth failure status | `401` | `403` | ## Model IDs Sarvam has one ID per model. Fireworks' Fast and US router variants, and your own dedicated deployments of these models, map to the same Sarvam ID. | Fireworks `model` | Sarvam `model` | Status on Sarvam | | ---------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------- | ---------------- | | `accounts/fireworks/models/glm-5p3` (also `routers/glm-5p3-fast`, `routers/glm-5p3-us`) | [`glm5.3`](/api/token-factory/models/glm-5-3) | GA | | `accounts/fireworks/models/glm-5p3-flash` (also `routers/glm-5p3-flash-us`) | [`glm5.3-flash`](/api/token-factory/models/glm-5-3-flash) | GA | | `accounts/fireworks/models/gemma-4-31b-it` (on-demand deployment on Fireworks) | [`gemma4`](/api/token-factory/models/gemma-4-31b) | GA | | `accounts/fireworks/models/deepseek-v4p1-flash` (also `routers/deepseek-v4p1-flash-us`) | [`deepseekv4.1-flash-1017`](/api/token-factory/models/deepseek-v4.1-flash-1017) | Beta | | `accounts/fireworks/routers/deepseek-v4p1-flash-1017-us` (US-only router on Fireworks) | [`deepseekv4.1-flash-1017`](/api/token-factory/models/deepseek-v4.1-flash-1017) | Beta | | `accounts/fireworks/models/deepseek-v4-pro` (April 2026 snapshot; on-demand deployment on Fireworks) | [`deepseekv4-pro`](/api/token-factory/models/deepseek-v4-pro) (0813 snapshot) | Beta | | `accounts/fireworks/models/kimi-k3` (also `routers/kimi-k3-fast`, `routers/kimi-k3-us`) | [`kimi-k3`](/api/token-factory/models/kimi-k3) | Beta | | Not offered on Fireworks | [`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash) | Beta | | Not offered on Fireworks (only the MiMo V2.6 Pro-RL variant is listed) | [`mimo-v2.6-pro`](/api/token-factory/models/mimo-v2.6-pro) | Beta | Sarvam's `deepseekv4-pro` is the newer 0813 snapshot of DeepSeek V4 Pro, so outputs can differ from the April 2026 snapshot on Fireworks. Gemma 4 31B and DeepSeek V4 Pro need a dedicated deployment on Fireworks. On Sarvam, call them by model ID like any other model, with no deployment to create or manage. Model IDs are case-sensitive. An unknown ID returns `404 not_found_error`. [`GET /v2/models`](/api-reference/chat/models-v2) lists the available models. Beta models need per-key access: without it, they return `400 invalid_request_error`, even though the list shows them. [Contact us](/api/getting-started/help) to request access. ## Models not on Sarvam If you use a model that Sarvam doesn't serve, start with the closest one below, then test it on your own prompts before switching traffic. These are suggestions, not drop-in equivalents. | If you use | Try on Sarvam | Why | | ----------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- | | GLM-5.3 Prime | [`glm5.3`](/api/token-factory/models/glm-5-3) | Same GLM-5.3 generation. | | GLM-5.3 FlashX | [`glm5.3-flash`](/api/token-factory/models/glm-5-3-flash) | Same fast GLM-5.3 variant family, with image input. | | Earlier GLM (GLM-5.2, GLM-5.1, GLM-5, GLM-4.x) | [`glm5.3`](/api/token-factory/models/glm-5-3) | Newer generation of the same family. For image input, use `glm5.3-flash`. | | DeepSeek V4 Flash (any snapshot), DeepSeek V3.x | [`deepseekv4.1-flash-1017`](/api/token-factory/models/deepseek-v4.1-flash-1017) | The DeepSeek Flash model Sarvam serves. | | DeepSeek V4 Pro (April 2026 snapshot), DeepSeek R1 | [`deepseekv4-pro`](/api/token-factory/models/deepseek-v4-pro) | Sarvam serves the newer 0813 snapshot. | | Kimi K2, K2.5, K2.6, K2.7 Code, K2 Thinking | [`kimi-k3`](/api/token-factory/models/kimi-k3) | Newer Kimi generation, with a longer 1M context. | | MiMo V2.5, MiMo V2.6 Pro UltraSpeed, MiMo V2.6 Pro-RL | [`mimo-v2.6-pro`](/api/token-factory/models/mimo-v2.6-pro) or [`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash) | Current MiMo generation, with image, audio, and video input. | | Gemma 3, Gemma 4 26B | [`gemma4`](/api/token-factory/models/gemma-4-31b) | Gemma 4 31B: image input, a 256K context window, and reasoning (off by default). | For other model families, such as Qwen, Llama, gpt-oss, or MiniMax, choose by what you need: | If you need | Try on Sarvam | | ---------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Indian-language quality | [`sarvam-105b`](/api/api-guides-tutorials/models/sarvam-105b) | | The lowest cost | [`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash) | | Image input | [`glm5.3-flash`](/api/token-factory/models/glm-5-3-flash), [`kimi-k3`](/api/token-factory/models/kimi-k3), [`gemma4`](/api/token-factory/models/gemma-4-31b), or the MiMo models | | Audio input | [`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash) or [`mimo-v2.6-pro`](/api/token-factory/models/mimo-v2.6-pro) | | Video input | [`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash) or [`mimo-v2.6-pro`](/api/token-factory/models/mimo-v2.6-pro) (chat only) | | The hardest coding and agent tasks | [`glm5.3`](/api/token-factory/models/glm-5-3) or [`deepseekv4-pro`](/api/token-factory/models/deepseek-v4-pro) | | Fast answers without reasoning | [`mimo-v2.6-flash`](/api/token-factory/models/mimo-v2.6-flash) with thinking off, or [`gemma4`](/api/token-factory/models/gemma-4-31b) | ## Media input and request size Sarvam accepts images, audio, and video only as **inline base64 data URIs**. Remote `http(s)` URLs return `400` on every model, even where your current provider fetches them for you, so download the file and encode it before sending. | Media | Models | Chat Completions | Responses | | ----- | ----------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------- | ----------------------------------- | | Image | Models with image input, such as `gemma4`, `glm5.3-flash`, `kimi-k3`, and the MiMo models | `image_url` part | `input_image` part | | Audio | `mimo-v2.6-flash`, `mimo-v2.6-pro` | `input_audio` part (`wav` or `mp3`), or an `audio_url` part with a data URI (WAV, MP3, FLAC, OGG, Opus) | `input_audio` part (`wav` or `mp3`) | | Video | `mimo-v2.6-flash`, `mimo-v2.6-pro`, and `kimi-k3` (beta, short clips) | `video_url` part | Not supported | ```python import base64 def data_uri(path: str, mime: str) -> str: with open(path, "rb") as f: return f"data:{mime};base64," + base64.b64encode(f.read()).decode() image_part = {"type": "image_url", "image_url": {"url": data_uri("chart.png", "image/png")}} video_part = {"type": "video_url", "video_url": {"url": data_uri("clip.mp4", "video/mp4")}} ``` Request bodies, including base64 media, are limited to **40 MiB** on `kimi-k3` and the MiMo models, and **10 MiB** on other models. Larger requests return `413` with the limit in the message. Base64 adds about 33%, so about 30 MB of raw media fits in a 40 MiB request. ## Errors and rate limits Sarvam returns `403` (not `401`) for a missing or invalid key, `429` when you exceed a rate limit, and `503` when a model is overloaded. Retry `429` and `503` with backoff. Limits apply per API key, as requests per minute rather than Fireworks' adaptive tokens-per-minute limits. See [Errors & Troubleshooting](/api/getting-started/errors-troubleshooting) and [Credits & Rate Limits](/api/getting-started/pricing#rate-limits). ## See also * [Migrating Chat Completion from Fireworks](/api/token-factory/migrations/from-fireworks/chat-completion) * [Migrating the Responses API from Fireworks](/api/token-factory/migrations/from-fireworks/responses-api) * [Open-Weight Models](/api/token-factory/models) * [Beta models](/api-reference/beta-apis) * [Errors & Troubleshooting](/api/getting-started/errors-troubleshooting)