pvakati-tech.ai

// ranked · October 2026

Best LLM for Browser Agents

Ranked by LiveBench’s browser category: navigating live sites to aggregate data, plan multi-step actions and act on what it finds within explicit constraints.

OpenRouter + LiveBenchAll rankingsAll models

GPT-6.1 Sol (93.3) and GPT-6.1 Terra (93.1) are effectively tied on LiveBench browser — less than a point apart, which effort settings alone can move. GPT-6 Astra follows at 91.3.

GPT-6.1 Sol vs GPT-6.1 Terra, head to head

Top 10

#ModelBrowser scoreBlended / 1MContextCapabilities
1GPT-6.1 Sol

openai/gpt-6.1-sol

93.3$4.001.1M
reasoningtoolsvision
2GPT-6.1 Terra

openai/gpt-6.1-terra

93.1$2.001.1M
reasoningtoolsvision
3GPT-6 Astra

openai/gpt-6-astra

91.3$20.001.1M
reasoningtoolsvision
4Claude Fable 5.1

anthropic/claude-fable-5.1

89.8$20.001M
reasoningtoolsvision
5Claude Sonnet 5.5

anthropic/claude-sonnet-5.5

89.3$4.001M
reasoningtoolsvision
6Claude Opus 5.5

anthropic/claude-opus-5.5

88.4$8.001M
reasoningtoolsvision
7Gemini 4 Argon

google/gemini-4-argon

82.8$4.001.0M
reasoningtoolsvision
8MiniMax M3.1 Flash

minimax/minimax-m3.1-flash

80.6$0.2621.0M
reasoningtoolsvision
9MiMo-V2.6-Pro

xiaomi/mimo-v2.6-pro

77.3$0.5441.1M
reasoningtoolsvision
10Kimi K3.1 Pro

moonshotai/kimi-k3.1-pro

75.9$4.001.0M
reasoningtoolsvision

18 more models qualify — see the full leaderboard.

FAQ

What is the best LLM for Browser Agents in October 2026?

GPT-6.1 Sol (93.3) and GPT-6.1 Terra (93.1) are effectively tied on LiveBench browser — less than a point apart, which effort settings alone can move. GPT-6 Astra follows at 91.3.

What is the best value in the top 10 for browser?

MiniMax M3.1 Flash. Lowest measured cost per point among the leaders — $0.0039 per point for a score of 80.6.

What is the best under $1.00 / 1m for browser?

MiniMax M3.1 Flash. Highest score at a blended list price of $1.00 per 1M tokens or less — 80.6 at $0.262.

How is this ranking produced?

Ranked by LiveBench’s browser category: navigating live sites to aggregate data, plan multi-step actions and act on what it finds within explicit constraints.

Other rankings

How this ranking is produced

  • One metric, stated above — nothing here is weighted or scored by us. Scores come from LiveBench release 2026-10-25; models without a published run don't appear in score-based rankings.
  • Price, context and capabilities — live from OpenRouter, refreshed every 15 minutes.
  • Ties — scores less than a point apart are called a tie; effort settings alone move a LiveBench score by more than that.

A public benchmark is someone else's workload. Before committing, see LLM & agent evaluation.