Using Dolphin 2.9.2 Qwen2-72B on Fireworks AI
Implementation guide · Dolphin · Cognitive Computations
Fireworks AI exposes Dolphin 2.9.2 Qwen2-72B through model ID accounts/fireworks/models/dolphin-2-9-2-qwen2-72b. Use the setup steps, sourced pricing, capabilities, and official provider links below to validate this route before deployment.
Last refreshed 2026-05-01. Next refresh: weekly.
Quick Start
- 1
- 2Use the Fireworks AI SDK or REST API to call
accounts/fireworks/models/dolphin-2-9-2-qwen2-72b— see the documentation for request format. - 3
Code Examples
pip install openaiFIREWORKS_API_KEYaccounts/fireworks/models/dolphin-2-9-2-qwen2-72bFireworks model IDs use "accounts/fireworks/models/{model-name}" format, e.g. "accounts/fireworks/models/llama4-scout-instruct-basic" or "accounts/fireworks/models/deepseek-r1".
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["FIREWORKS_API_KEY"],
base_url="https://api.fireworks.ai/inference/v1"
)
response = client.chat.completions.create(
model="accounts/fireworks/models/dolphin-2-9-2-qwen2-72b",
messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)Pricing on Fireworks AI
| Type | Price (per 1M) |
|---|---|
| Input tokens | $0.90 |
| Output tokens | $0.90 |
Capabilities
About Dolphin 2.9.2 Qwen2-72B
Dolphin 2.9.2 Qwen2-72B is Cognitive Computations' uncensored full-weight fine-tune of Qwen2-72B. The model card documents ChatML formatting, initial agentic abilities, and function calling support; it also notes the Qwen2 base has a 128K context window while the fine-tune used 8K training sequences. Because Dolphin removes much of the usual alignment filtering, deployments should add their own moderation and policy layer before exposing it as a public service.