Qwen 3

Qwen 3 on Fireworks AI: Controllable Chain-of-Thought and Tool Calling at Frontier Scale

GLM 5.2 is live! Opus-level intelligence at open-source rates. Pay per token on serverless. Try it today.

Blog

Qwen 3 Qwen 3 on Fireworks AI: Controllable Chain-of-Thought and Tool Calling at Frontier Scale

PUBLISHED 5/6/2025

Table of Contents TL;DR Why this release matters ? 15-second quick-start Under the hood: The Hybrid Thinking Switch Model card (Fireworks hosted) Performance tips Using the Fireworks endpoint Closing thoughts

Table of Contents

TL;DR

• Reasoning meets function calls. Qwen 3 now streams an explicit … trace and the exact JSON tool call in the same completion. • Turbo or stealth—your choice. Flip reasoning_effort="none" (or use the /think / /no_think tags) to trade transparency for raw throughput on the fly. • Mixture-of-Experts giant, pay-as-you-go. The 235 B-parameter / 22 B-active Qwen3-235B-A22B runs serverlessly on Fireworks. • Drop-in OpenAI compatibility. Use the Fireworks endpoint with the official OpenAI client; everything else stays the same.

Why this release matters ?

Until now, open-source LLMs forced a choice: show the chain of thought or call tools deterministically . Qwen 3’s new architecture does both in one pass, and keeps the reasoning block segregated so downstream code can ignore or audit it at will. Pair that with a 128-expert MoE that only activates eight experts (≈22 B live parameters) and you get near-frontier quality at a fraction of the compute- fully Apache-2.0 and live on Fireworks today ( Fireworks - Qwen3 235B-A22B model ). 15-second quick-start

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 from openai import OpenAI import os , json

client = OpenAI ( base_url = "" , api_key = os . environ [ "FIREWORKS_API_KEY" ] , )

messages = [ { "role" : "user" , "content" : "What’s the weather in Boston today?" } ]

tools = [ { "type" : "function" , "function" : { "name" : "get_weather" , "description" : "Return current weather for a US city" , "parameters" : { "type" : "object" , "properties" : { "location" : { "type" : "string" } } , "required" : [ "location" ] } } } ]

Reasoning based tool calls

resp = client . chat . completions . create ( model = "accounts/fireworks/models/qwen3-235b-a22b" , messages = messages , tools = tools , max_tokens = 4096 , temperature = 0.6 , )

first = resp . choices [ 0 ] . message print ( first . content ) # contains … print ( first . tool_calls )

Non - reasoning based tool calls

resp = client . chat . completions . create ( model = "accounts/fireworks/models/qwen3-235b-a22b" , messages = messages , tools = tools , max_tokens = 4096 , temperature = 0.6 , extra_body = { "reasoning_effort" : "none" , } , )

second = resp . choices [ 0 ] . message print ( second . content ) # does not contain … print ( second . tool_calls )

The first call contains reasoning chain of thought + tool call, the second doesn’t think, and just makes the tool calls. Under the hood: The Hybrid Thinking Switch

• Thinking mode ( reasoning_effort!=”none” ) • Generates … and a final answer. • Recommended params: temperature ≈ 0.6, top_p ≈ 0.95, top_k = 20 .

• Non-thinking mode ( reasoning_effort=”none” or a /no_think tag) • Omits the reasoning block to save tokens and latency. • Use slightly spicier sampling: temperature ≈ 0.7, top_p ≈ 0.8 .

Because the trace sits in its own tag, you can log, redact, or meter it independently- the same pattern we covered in Constrained Generation with Reasoning . Model card (Fireworks hosted)

• Qwen 3-235B-A22B

Total parameters 235 B (Mixture-of-Experts) Active parameters 22 B (8 / 128 experts) Layers 94 Attention heads 64 / 4 Context window 32 768 tokens (native) 131 072 with YaRN License Apache-2.0 Endpoint accounts/fireworks/models/qwen3-235b-a22b

Performance tips

• Long answers – Allocate at least 4 k output tokens for essays; up to 32 k for book-length generations. • Cost & speed control – Invoke reasoning only on the turns that need it, then strip before storage.

Using the Fireworks endpoint

Our endpoint is fully OpenAI compatible, please give it a try! 1 2 3 4 5 6 7 8 curl https : / / api . fireworks . ai / inference / v1 / chat / completions \

• H "Authorization: Bearer $FIREWORKS_API_KEY" \
• H "Content-Type: application/json" \
• d ' {

"model" : "accounts/fireworks/models/qwen3-235b-a22b" , "messages" : [ { "role" : "user" , "content" : "Translate 这本书多少钱？" } ] , "reasoning_effort" : "none" } '

Closing thoughts

With Qwen 3-235B-A22B , open-source finally gets a model that: Reveals its chain of thought when you ask. Emits tool calls in the exact same request. Scales to frontier-size contexts- all under Apache-2.0.

No secret weights, no bespoke SDKs. Just point your existing OpenAI-style client at Fireworks and build. Questions, feedback, or cool demos? Drop by our Discord or tag us on X. Happy shipping!