Complete list of supported AI models, exact model IDs for copy-pasting, context lengths, and token pricing.

Supported AI Models & Model IDs

Fusion AI Gateway provides high-throughput, unquantized frontier models through a single OpenAI-compatible /v1/chat/completions endpoint.

Copy and paste the exact Model IDs below into your API requests, CLI configurations, or agent tools.

MODEL	MODE	EXACT MODEL ID (Copy & Paste)	CONTEXT	INPUT / 1M	CACHED	OUTPUT / 1M	SPEED (TPS)	TTFT
DeepSeek-V4-Flash `POPULAR`	🚀 Fast	`deepseek-ai/DeepSeek-V4-Flash-0731`	1M / 384K	$0.10	$0.014	$0.18	~257 t/s (cached ~45K t/s)	~448 ms (cached ~55 ms)
DeepSeek-V4-Flash	⚡ Think	`deepseek-ai/DeepSeek-V4-Flash-0731`	1M / 384K	$0.10	$0.014	$0.18	~540 t/s (cached ~45K t/s)	~3.2 s (cached ~66 ms)
MiniMax-M2.7	—	`MiniMaxAI/MiniMax-M2.7`	204K / 131K	$0.25	$0.02	$1.00	~117 t/s (cached ~82K t/s)	~272 ms (cached ~69 ms)
GLM-5.2 `SOON`	—	Coming Soon	128K	TBD	TBD	TBD	Soon	Soon

2. Model Deep Dive & Parameter Flags

1. DeepSeek V4 Flash (07-31)

Model ID: deepseek-ai/DeepSeek-V4-Flash-0731
Context Length: 1,048,576 tokens (1M context)
Precision: Full FP8 unquantized MoE
Features: 1M context window, fast tool-calling, optional deep reasoning mode.

Mode Selection:

Fast / Direct Mode (Default): Set enable_thinking: false or omit for sub-400ms TTFT and ~85 tokens/sec generation speed.
Reasoning Mode: Pass enable_thinking: true to generate full step-by-step thinking traces in delta.reasoning.

{
  "model": "deepseek-ai/DeepSeek-V4-Flash-0731",
  "messages": [{"role": "user", "content": "Explain how distributed cache locking works."}],
  "enable_thinking": false,
  "stream": true
}

2. MiniMax M2.7

Model ID: MiniMaxAI/MiniMax-M2.7
Context Length: 204,800 tokens (204K context)
Precision: Full FP8 MoE
Features: Lowest TTFT on the network (~270ms), ultra-fast code generation, robust multi-tool orchestration.

{
  "model": "MiniMaxAI/MiniMax-M2.7",
  "messages": [{"role": "user", "content": "Refactor this TypeScript function for maximum throughput."}],
  "stream": true
}

3. Example Code Snippets

cURL Example

curl -X POST https://api.fusioncode.app/v1/chat/completions \
  -H "Authorization: Bearer fc_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-ai/DeepSeek-V4-Flash-0731",
    "messages": [
      {"role": "system", "content": "You are a senior software architect."},
      {"role": "user", "content": "How do you handle cache invalidation across edge nodes?"}
    ],
    "stream": true
  }'

Python (OpenAI SDK)

from openai import OpenAI

client = OpenAI(
    api_key="fc_your_api_key_here",
    base_url="https://api.fusioncode.app/v1"
)

response = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V4-Flash-0731",
    messages=[
        {"role": "user", "content": "Hello Fusion Gateway!"}
    ],
    stream=True
)

for chunk in response:
    content = chunk.choices[0].delta.content or ""
    print(content, end="", flush=True)

TypeScript / Node (OpenAI SDK)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.FUSION_API_KEY || "fc_your_api_key_here",
  baseURL: "https://api.fusioncode.app/v1",
});

async function main() {
  const stream = await client.chat.completions.create({
    model: "deepseek-ai/DeepSeek-V4-Flash-0731",
    messages: [{ role: "user", content: "Explain async iterators in TypeScript." }],
    stream: true,
  });

  for await (const chunk of stream) {
    process.stdout.write(chunk.choices[0]?.delta?.content || "");
  }
}

main();

4. Querying Models Programmatically

You can list all currently active models and their token pricing live from the Gateway:

curl -X GET https://api.fusioncode.app/v1/models \
  -H "Authorization: Bearer fc_your_api_key_here"

Example Response:

{
  "object": "list",
  "data": [
    {
      "id": "deepseek-ai/DeepSeek-V4-Flash-0731",
      "object": "model",
      "display_name": "DeepSeek V4 Flash",
      "context_length": 1048576,
      "pricing": {
        "input": 0.10,
        "cached": 0.014,
        "output": 0.18
      }
    },
    {
      "id": "MiniMaxAI/MiniMax-M2.7",
      "object": "model",
      "display_name": "MiniMax M2.7",
      "context_length": 204800,
      "pricing": {
        "input": 0.25,
        "cached": 0.02,
        "output": 1.00
      }
    }
  ]
}