> 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 the Responses API from Fireworks AI to Sarvam > Moving from Fireworks' stored Responses API to Sarvam's stateless /v2/responses: conversation history, reasoning, function tools, server-side tools, and fields to change. *Last updated: October 23, 2026* This guide covers moving Fireworks AI [Responses API](https://docs.fireworks.ai/guides/response-api) calls to Sarvam's [`POST /v2/responses`](/api/token-factory/responses-api). For Chat Completions, see [Migrating Chat Completion](/api/token-factory/migrations/from-fireworks/chat-completion). > **Tip** > > **In short:** Fireworks stores responses and lets you chain calls with > `previous_response_id`. Sarvam's Responses API is **stateless**: keep the conversation in > your app and send it in `input` on every call. Change `base_url`, `api_key`, and `model`, and > replace server-side tools (`web_search`, `mcp`, `python`) with function tools you run yourself. ## Migration checklist 1. Change `base_url` to `https://api.sarvam.ai/v2` and `api_key` to your Sarvam key. 2. Replace each model ID using the [model ID table](/api/token-factory/migrations/from-fireworks/overview#model-ids). 3. Replace `previous_response_id` chaining with a history list sent in `input` ([Conversation history](#conversation-history)). 4. Remove `store`, `background`, and calls to retrieve stored responses ([list](#fields-to-remove)). 5. Replace server-side tools with function tools ([Tools](#tools)). 6. On `glm5.3` and `glm5.3-flash`, don't send `"max"`; omitting `reasoning.effort` gives `max` on `glm5.3` and `high` on `glm5.3-flash` ([Reasoning](#reasoning)). 7. Set `max_output_tokens`: it has no default cap on Sarvam. ## Quick start #### Before (Fireworks) #### Python (OpenAI SDK) ```python from openai import OpenAI client = OpenAI( api_key="YOUR_FIREWORKS_API_KEY", base_url="https://api.fireworks.ai/inference/v1", ) response = client.responses.create( model="accounts/fireworks/models/kimi-k3", instructions="You are a helpful assistant.", input="Explain database indexing in three short points.", reasoning={"effort": "low"}, max_output_tokens=3000, ) print(response.output_text) ``` #### JavaScript (OpenAI SDK) ```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.responses.create({ model: "accounts/fireworks/models/kimi-k3", instructions: "You are a helpful assistant.", input: "Explain database indexing in three short points.", reasoning: { effort: "low" }, max_output_tokens: 3000, }); console.log(response.output_text); ``` #### cURL ```bash curl https://api.fireworks.ai/inference/v1/responses \ -H "Authorization: Bearer YOUR_FIREWORKS_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "accounts/fireworks/models/kimi-k3", "input": "Explain database indexing in three short points.", "reasoning": {"effort": "low"}, "max_output_tokens": 3000 }' ``` #### After (Sarvam) #### Python (OpenAI SDK) ```python from openai import OpenAI client = OpenAI( api_key="YOUR_SARVAM_API_KEY", base_url="https://api.sarvam.ai/v2", ) response = client.responses.create( model="kimi-k3", instructions="You are a helpful assistant.", input="Explain database indexing in three short points.", reasoning={"effort": "low"}, max_output_tokens=3000, ) print(response.output_text) ``` #### JavaScript (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.responses.create({ model: "kimi-k3", instructions: "You are a helpful assistant.", input: "Explain database indexing in three short points.", reasoning: { effort: "low" }, max_output_tokens: 3000, }); console.log(response.output_text); ``` #### Python (Sarvam SDK) ```python from sarvamai import SarvamAI client = SarvamAI(api_subscription_key="YOUR_SARVAM_API_KEY") response = client.responses.create( model="kimi-k3", instructions="You are a helpful assistant.", input="Explain database indexing in three short points.", reasoning={"effort": "low"}, max_output_tokens=3000, ) # The Sarvam SDK has no output_text shortcut; read the message item. message = next(item for item in response.output if item.type == "message") print(message.content[0].text) ``` #### JavaScript (Sarvam SDK) ```javascript import { SarvamAIClient } from "sarvamai"; const client = new SarvamAIClient({ apiSubscriptionKey: "YOUR_SARVAM_API_KEY" }); const response = await client.responses.create({ model: "kimi-k3", instructions: "You are a helpful assistant.", input: "Explain database indexing in three short points.", reasoning: { effort: "low" }, max_output_tokens: 3000, }); const message = response.output.find((item) => item.type === "message"); console.log(message.content[0].text); ``` #### cURL ```bash curl https://api.sarvam.ai/v2/responses \ -H "api-subscription-key: YOUR_SARVAM_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "kimi-k3", "input": "Explain database indexing in three short points.", "reasoning": {"effort": "low"}, "max_output_tokens": 3000 }' ``` ## Conversation history On Fireworks, each response is stored and the next call can point at it with `previous_response_id`. On Sarvam, nothing is stored: keep the turns in a list and send the whole list in `input`. `previous_response_id`, `store: true`, `conversation`, and `background: true` return `400`, and `GET /v2/responses/{id}` returns `404`. #### Before (Fireworks) ```python from openai import OpenAI client = OpenAI( api_key="YOUR_FIREWORKS_API_KEY", base_url="https://api.fireworks.ai/inference/v1", ) first = client.responses.create( model="accounts/fireworks/models/glm-5p3", input="My name is Asha.", ) second = client.responses.create( model="accounts/fireworks/models/glm-5p3", input="What's my name?", previous_response_id=first.id, ) print(second.output_text) ``` #### After (OpenAI SDK) ```python from openai import OpenAI client = OpenAI( api_key="YOUR_SARVAM_API_KEY", base_url="https://api.sarvam.ai/v2", ) history = [{"role": "user", "content": "My name is Asha."}] first = client.responses.create(model="glm5.3", input=history) history += [ {"role": "assistant", "content": first.output_text}, {"role": "user", "content": "What's my name?"}, ] second = client.responses.create(model="glm5.3", input=history) print(second.output_text) ``` #### After (Sarvam SDK) ```python from sarvamai import SarvamAI client = SarvamAI(api_subscription_key="YOUR_SARVAM_API_KEY") def reply_text(response): message = next(item for item in response.output if item.type == "message") return message.content[0].text history = [{"role": "user", "content": "My name is Asha."}] first = client.responses.create(model="glm5.3", input=history) history += [ {"role": "assistant", "content": reply_text(first)}, {"role": "user", "content": "What's my name?"}, ] second = client.responses.create(model="glm5.3", input=history) print(reply_text(second)) ``` Long conversations grow `input` on every turn. Trim or summarize older turns yourself to stay within the model's context window. ## Reasoning | Fireworks | Sarvam | Note | | ----------------------------------------- | --------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- | | `reasoning.effort` (`low`, `high`, `max`) | Same | On `glm5.3` and `glm5.3-flash`, `"max"` returns `400`; omitting the field gives `max` on `glm5.3` and `high` on `glm5.3-flash`. | | `reasoning.effort: "medium"` / `"xhigh"` | Use `"low"`, `"high"`, or `"max"` | Accepted on some models and returns `400` on others. | | `reasoning.effort: "none"` | Same | Turns thinking off on `deepseekv4-pro`, `mimo-v2.6-flash`, and `mimo-v2.6-pro`. | | Reasoning item in `output` | Same | Read the trace from the item's `content[0].text`. `summary` is filled only when you set `reasoning.summary`. | Reasoning tokens count toward `max_output_tokens`, which has no default cap on Sarvam. ## Tools Sarvam's Responses API supports **function tools** only, in the flat Responses format: `{"type": "function", "name": ..., "parameters": ...}`. | Fireworks tool | On Sarvam | | -------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `function` | Same | | `web_search` | Returns `400` on most models; `gemma4` and `sarvam-105b` accept it and never call it. Call a search service yourself and expose it as a function tool. | | `mcp`, `sse` | Returns `400` on most models; `gemma4` and `sarvam-105b` accept it and never call it. Call your MCP server from your own code and expose its tools as function tools. | | `python` | Returns `400`. Run code in your own sandbox and return the result as a function call output. | The function call round trip is the same as on Fireworks, except that you resend the history instead of passing `previous_response_id`: send the `function_call` item and a `function_call_output` item in the next request's `input`. See the [OpenRouter Responses guide's full example](/api/token-factory/migrations/from-openrouter/responses-api#full-example), which works the same way for any stateless client. `tool_choice` accepts `"auto"`, `"none"`, `"required"`, or a named function. `parallel_tool_calls: false` limits a turn to one tool call on every open-weight model, and `max_tool_calls` returns `400`. ## Fields to remove | Fireworks field or call | What to do on Sarvam | | ---------------------------------------------------------------------------------- | ------------------------------------------------------------------------- | | `previous_response_id`, `conversation` | Send the full history in `input`. | | `store` | Remove, or set it to `false`. `true` returns `400`. | | `background` | Remove. Make a normal (or streaming) request. | | `GET /responses/{id}`, `GET /responses/{id}/input_items`, `DELETE /responses/{id}` | Not available. Keep responses in your own storage if you need them later. | | `stop` | Remove. Not a Responses parameter on Sarvam. | | `max_tool_calls` | Remove. Stop the tool loop in your own code. | ## Parameter concept map | Fireworks | Sarvam | Note | | -------------------------------------------- | ------------------------ | --------------------------------------------------------------------------------------------------- | | `input` (string or items), `instructions` | Same | | | `max_output_tokens`, `temperature`, `top_p` | Same | Set `max_output_tokens`: there's no default cap. | | `text.format` (`json_schema`, `json_object`) | Same | | | `input_image` content parts | Same | Inline base64 data URIs only, on models with image input. | | `input_audio` content parts | Same, on the MiMo models | `wav` or `mp3` only on the Responses API. | | Video input | Not supported | Send video on Chat Completions. | | `stream: true` | Same | Standard events: `response.created`, `response.output_text.delta`, `response.completed`, and so on. | | `response.output_text` (OpenAI SDK) | Same | The `sarvamai` SDK doesn't have this shortcut; read the `message` item in `output`. | ## Common mistakes #### Passing previous_response_id Sarvam returns `400`. Keep the history in your app and send it in `input`, as in [Conversation history](#conversation-history). #### Fetching a stored response by id Nothing is stored, so `GET /v2/responses/{id}` returns `404`. Save what you need from each response when you receive it. #### Using server-side tools `web_search`, `mcp`, and `python` tools aren't run on Sarvam. Most models return `400`; `gemma4` and `sarvam-105b` accept `web_search` and `mcp` but never call them. Run them yourself and expose them as function tools. #### Sending reasoning.effort: max to glm5.3 `glm5.3` and `glm5.3-flash` return `400` for the literal string `"max"`. Omit the field: that gives `max` on `glm5.3` and `high` on `glm5.3-flash`. ## See also * [Migration overview](/api/token-factory/migrations/from-fireworks/overview): base URL, auth, and model IDs * [Migrating Chat Completion from Fireworks](/api/token-factory/migrations/from-fireworks/chat-completion) * [Responses API guide](/api/token-factory/responses-api) * [Responses API reference](/api-reference/responses/create)