Models & Pricing

Vyre integrates with the world's leading AI models. Switch between models mid-conversation or set a default. Each model has different strengths — use the right model for the task.

Available models

NameDefault ContextMax ContextCapabilities
GPT-5.6 Sol
1M1M
GPT-5.6 Terra
1M1M
GPT-5.6 Luna
1M1M
GPT-5.4
272k272k
GPT-5.4 Mini
272k272k
GPT-5.4 Nano
272k272k
GPT-5.3 Codex
272k272k
GPT-5.2
400k400k
GPT-5.2 Codex
400k400k
GPT-5.1
400k400k
GPT-5.1 Codex
400k400k
GPT-5.1 Codex Mini
400k400k
GPT-5
400k400k
GPT-5 Mini
400k400k
GPT-5 Nano
400k400k
GPT-5 Codex
400k400k
GPT-5 Pro
400k400k
GPT-4.1
1M1M
GPT-4.1 Mini
1M1M
GPT-4.1 Nano
1M1M
o4-mini
200k200k
Claude Sonnet 4.6
1M1M
Claude Opus 4.8
1M1M
Claude Opus 4.7
1M1M
Claude Opus 4.6
1M1M
Claude Opus 4.5
200k200k
Claude Sonnet 4.5
200k200k
Claude Haiku 4.5
200k200k
Gemini 3.5 Flash
1M1M
Gemini 3.1 Pro Preview
1M1M
Gemini 3.1 Flash Lite
1M1M
Gemini 3 Flash (Preview)
1M1M
Gemini 2.5 Pro
1M1M
Gemini 2.5 Flash
1M1M
Gemini 2.5 Flash Lite
1M1M

Choosing a model

For everyday coding tasks

Use Gemini 3.5 Flash, GPT-5.6 Sol, or Claude Sonnet 4.6 as your default. They handle the vast majority of coding tasks with top-tier speed and reasoning.

For hard architectural problems

Switch to Claude Opus 4.8, GPT-5.5 Pro, or Gemini 3.1 Pro Preview when you need deep reasoning — complex migrations, system design decisions, or subtle concurrency bugs.

For quick edits and rapid exploration

Use GPT-5.4 Nano, GPT-4.1 Nano, Gemini 3.1 Flash Lite, or Claude Haiku 4.5 when you need instant responses.

For massive codebases

Use Gemini 3.5 Flash, GPT-4.1, or Claude Sonnet 4.6 to take advantage of 1M+ token context windows.

Pricing

Plan Included requests Additional requests
Free 50 requests/month Not available
Pro 500 requests/month $0.02 per request
Business 2,000 requests/month $0.015 per request
Enterprise Unlimited Custom pricing

A "request" is one complete agent turn — from your message to Agent's response. Tool calls within a single turn don't count as separate requests.

Usage limits

View your current usage in Settings → Billing → Usage. You'll see:

  • Requests used this billing period
  • Requests remaining
  • Breakdown by model
  • Estimated cost for the current period

When you're at 80% of your plan's included requests, Vyre shows a warning banner. At 100%, requests using expensive models (Opus, GPT-5 Pro) are paused until the next billing cycle or until you upgrade. Fast models continue to work for quick tasks.