AI Models - SynapseX

Overview

SynapseX provides access to the world’s leading AI models through a unified API endpoint. All models are available via the SynapseX inference platform — no need to manage multiple API keys or provider accounts. Switch between models instantly in the VS Code Agent or configure your preferred default model in the CLI.

Recommended Models

Claude Sonnet 4

Best for: Complex reasoning, architecture design, code reviewAnthropic’s flagship model for enterprise coding. Exceptional at multi-step reasoning and understanding large codebases.Model ID: claude-sonnet-4-5

GPT-4o

Best for: Rapid code generation, multimodal tasksOpenAI’s flagship model with vision capabilities. Excellent for UI generation from screenshots and fast iteration.Model ID: gpt-4o

Gemini 2.5 Pro

Best for: Long context, entire-codebase analysisGoogle’s frontier model with the largest context window. Ideal for analyzing large repositories or long documents.Model ID: gemini-2.5-pro

DeepSeek R1

Best for: Mathematical reasoning, algorithm designState-of-the-art open-source model with chain-of-thought reasoning. Cost-effective for compute-intensive tasks.Model ID: deepseek-r1

Full Model Reference

Anthropic — Claude Models

Model	Context	Best For
`claude-opus-4`	200K tokens	Hardest tasks, deep reasoning, long-horizon planning
`claude-sonnet-4-5`	200K tokens	Balanced performance — recommended default
`claude-haiku-3-5`	200K tokens	Fast, cost-efficient for simple tasks
`claude-3-7-sonnet`	200K tokens	Extended thinking for complex math and logic

OpenAI — GPT Models

Model	Context	Best For
`gpt-4o`	128K tokens	Multimodal, vision, fast generation
`gpt-4o-mini`	128K tokens	Cost-effective for high-volume tasks
`o3`	200K tokens	Advanced reasoning with extended thinking
`o4-mini`	200K tokens	Fast reasoning for coding tasks
`gpt-4.1`	1M tokens	Very long context, document analysis

Google — Gemini Models

Model	Context	Best For
`gemini-2.5-pro`	1M tokens	Largest context, full repo analysis
`gemini-2.5-flash`	1M tokens	Fast multimodal processing
`gemini-2.0-flash`	1M tokens	Real-time streaming, low latency

Meta — Llama Models

Model	Context	Best For
`llama-4-maverick`	1M tokens	Open source, self-hosted option
`llama-4-scout`	10M tokens	Massive context window tasks
`llama-3.3-70b`	128K tokens	Cost-effective open source

DeepSeek Models

Model	Context	Best For
`deepseek-r1`	64K tokens	Chain-of-thought reasoning, math
`deepseek-v3`	64K tokens	High-quality general coding
`deepseek-r1-lite`	32K tokens	Cost-efficient reasoning

Mistral Models

Model	Context	Best For
`mistral-large-2`	128K tokens	European data residency compliance
`mixtral-8x22b`	64K tokens	High throughput tasks
`codestral`	256K tokens	Code-specific generation

Perplexity — Search-Augmented Models

Model	Context	Best For
`sonar-pro`	128K tokens	Web-augmented queries, real-time info
`sonar-reasoning`	128K tokens	Research with citations
`sonar`	128K tokens	Fast web search

Qwen Models (Alibaba Cloud)

Model	Context	Best For
`qwen2.5-72b`	128K tokens	Cost-effective at scale
`qwen2.5-coder-32b`	128K tokens	Specialized code generation
`qwq-32b`	32K tokens	Reasoning tasks

Model Selection Guide

By Task Type

Task	Recommended Model
Architecture design & planning	`claude-opus-4` or `claude-sonnet-4-5`
Rapid prototyping	`gpt-4o` or `claude-haiku-3-5`
Large codebase analysis	`gemini-2.5-pro`
Security audit	`claude-sonnet-4-5`
Math & algorithm design	`deepseek-r1` or `o3`
UI generation from screenshots	`gpt-4o` (vision)
Web-augmented research	`sonar-pro`
High-volume automation	`deepseek-v3` or `qwen2.5-72b`
EU data residency	`mistral-large-2`

By Context Size Needed

Context Required	Use Model
< 32K tokens (small files)	`claude-haiku-3-5`, `gpt-4o-mini`
32K–128K tokens (modules)	`claude-sonnet-4-5`, `gpt-4o`
128K–200K tokens (large features)	`claude-opus-4`, `gemini-2.5-flash`
200K–1M tokens (full repositories)	`gemini-2.5-pro`, `gpt-4.1`
> 1M tokens (massive codebases)	`llama-4-scout` (10M context)

Switching Models

In VS Code Agent

Click the model selector dropdown in the SynapseX chat panel. Your conversation context is preserved when switching models mid-session.

In SynapseX CLI

# Create an agent with a specific model
synapsex agent create repo-analyzer --model gpt-4o

# Invoke that agent
synapsex agent invoke repo-analyzer "Analyze this repository"

Via API

import requests

response = requests.post(
    "https://api.synapsex.ai/v1/chat/completions",
    headers={"Authorization": f"Bearer {SYNAPSEX_API_KEY}"},
    json={
        "model": "claude-sonnet-4-5",
        "messages": [{"role": "user", "content": "Explain this code..."}]
    }
)

Model Availability

All models are available through:

SynapseX VS Code Agent — model selector in chat panel
SynapseX CLI — synapsex agent create --model <model>
SynapseX API — model field in chat completions request
SynapseX Platform — model selector in the web dashboard

Model availability may vary by subscription plan. Contact support for enterprise access to additional models or higher rate limits.

API Overview

Getting started with the SynapseX API

Authentication

API keys and authentication methods

Best Practices

When to use which model

VS Code Agent

Configure your default model

​Overview

​Recommended Models

Claude Sonnet 4

GPT-4o

Gemini 2.5 Pro

DeepSeek R1

​Full Model Reference

​Anthropic — Claude Models

​OpenAI — GPT Models

​Google — Gemini Models

​Meta — Llama Models

​DeepSeek Models

​Mistral Models

​Perplexity — Search-Augmented Models

​Qwen Models (Alibaba Cloud)

​Model Selection Guide

​By Task Type

​By Context Size Needed

​Switching Models

​In VS Code Agent

​In SynapseX CLI

​Via API

​Model Availability

​Related Resources

API Overview

Authentication

Best Practices

VS Code Agent

Overview

Recommended Models

Full Model Reference

Anthropic — Claude Models

OpenAI — GPT Models

Google — Gemini Models

Meta — Llama Models

DeepSeek Models

Mistral Models

Perplexity — Search-Augmented Models

Qwen Models (Alibaba Cloud)

Model Selection Guide

By Task Type

By Context Size Needed

Switching Models

In VS Code Agent

In SynapseX CLI

Via API

Model Availability

Related Resources