Coding Agent Index: Sonnet 5.5 tops at 68 but costs $14.19/task vs GPT-6.1 Sol's $1.04
ArtificialAnlys · x · 2026-10-02
Artificial Analysis updated its Coding Agent Index (DeepSWE v1.1, Terminal-Bench 4.0, SWE-Atlas-QnA) with three new frontier models near the top:
- Claude Sonnet 5.5 (max) in Claude Code takes first place at 68, but with the highest measured cost at $14.19 per task
- Gemini 4 Argon (high) in Antigravity CLI scores 64 at $5.84/task — less than half of Sonnet's cost; note it uses Google's promotional pricing and Argon isn't publicly available yet
- GPT-6.1 Sol (xhigh) in Codex scores 63 at just $1.04/task, roughly one-sixth of Argon's cost
Takeaway: performance is tightly clustered while costs span an order of magnitude, making cost-per-task the key differentiator.
Related event: Coding Agent Index: Claude Leads at Far Higher Cost(2 posts)→
More from coding & agent
- Lightfoot: Exact, Model-Free Document Deletion for GraphRAG in 14 Seconds — JeremyCMorgan · 2026-11-20
- LangChain adds natural-language AI Search to LangSmith trace filtering — LangChain · 2026-11-20
- OpenAI Touts GPT-Live-1.1 for Building Voice Agents in Your Apps — OpenAIDevs · 2026-11-20
- LexiEngine: Rust inference engine for AMD runs 27B at full 262k context 47% faster than llama.cpp — W61k3r · 2026-11-20
- ENTIA launches Entity Verification MCP: 52M+ business records across 34 countries for AI agents — modelcontextprotocol · 2026-11-20
- Personal Agents Weekly: a roundup of the personal-agent ecosystem — heyneighbor · 2026-11-20