← comparisons · head to head
Claude Haiku 5.5 vs GPT-6.1 Sol
GPT-6.1 Sol scores higher than Claude Haiku 5.5 on all 6 public leaderboards that rate both. On the ArtificialWatch Index, which takes the median of each model's percentile across the boards that score it, Claude Haiku 5.5 is at 52 and GPT-6.1 Sol at 81.8 (out of 100). On each maker's API, Claude Haiku 5.5 costs $0.10 / $0.50 per 1M tokens (input / output) and GPT-6.1 Sol $2 / $10.
Scores as published by each board; table composed November 13, 2026.
Get the next launch alert, free →
Leaderboard by leaderboard
| Board | Claude Haiku 5.5 | GPT-6.1 Sol | Higher |
|---|---|---|---|
| LMArena Elo (community votes) | 1,593.4 | 1,743.3 | GPT-6.1 Sol |
| LiveBench global average (0-100) | 68.9 | 81.4 | GPT-6.1 Sol |
| SimpleBench AVG@5 % (private set) | 59.2 | 82.9 | GPT-6.1 Sol |
| Vals AI Vals Index accuracy % · weighted finance + coding tasks | 58.7 | 63.9 | GPT-6.1 Sol |
| Artificial Analysis intelligence index · measured runs only | 36.5 | 54.6 | GPT-6.1 Sol |
| Epoch AI Epoch Capabilities Index (ECI) · many benchmarks on one scale | 160.4 | 168.5 | GPT-6.1 Sol |
Raw scores are each board's own scale, so compare within a row, not across rows. Full table: the bench index.
For coding
On the ArtificialWatch coding index, Claude Haiku 5.5 is at 59.3 (#18 of 48) and GPT-6.1 Sol at 74.1 (#12). Claude Haiku 5.5 scores higher on 1 of the 3 coding boards that rate both and GPT-6.1 Sol on 2.
| Coding board | Claude Haiku 5.5 | GPT-6.1 Sol |
|---|---|---|
| LMArena · Coding | 1,453.6 | 1,485.6 |
| LiveBench · Coding | 77.6 | 82.1 |
| Vals Vibe Code Bench | 90.4 | 88.9 |
Claude Haiku 5.5 at a glance
- Developer
- Anthropic
- Status
- Live since October 7, 2026
- API price
- $0.10 input / $0.50 output per 1M tokens · Anthropic, claude-haiku-5-5
- Context window
- 1M tokens
Live — Oct 7 · $0.10/$0.50 per MTok for prompts up to 100K tokens, $0.50/$2.50 above · cache reads $0.01 / $0.05 · 1M ctx / 128K out · knowledge cutoff Jun 2026 · in the API, the Claude apps, AWS, Google Cloud and Microsoft Foundry the same day · Anthropic's own table: OSWorld 2.1 72.4% (GPT-6 Luna 48.9%, Haiku 4.5 15.7%), Terminal-Bench 4.0 39.2% (Luna 16.4%) · Artificial Analysis Intelligence Index 43 at max effort, #2 of 179 in its price tier (read Oct 7) · our sweep fired on claude-haiku-5-5 at 17:52 UTC, OpenRouter listed it at 18:31.
Claude Haiku 5.5 is out at $0.10 / $0.50, a tenth of Haiku 4.5's price
2026-10-07 · Anthropic released Claude Haiku 5.5 on October 7 as claude-haiku-5-5, on the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry and in the Claude apps.
GPT-6.1 Sol at a glance
- Developer
- OpenAI
- Status
- Live since September 29, 2026
- API price
- $2 input / $10 output per 1M tokens · OpenAI, gpt-6.1-sol
- Context window
- 1.05M tokens
Live — Sep 29, launched at DevDay · Ultrafast tier open to all API users Oct 8 at $12/$60 (six times standard, OpenAI's pricing page) · $2/$10 per 1M, cached input $0.10 (half of GPT-6 Sol's) · in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu, not yet in Chat · OpenAI says it nearly matches GPT-6 Astra on agentic coding, computer use and professional work at a fifth of Astra's price · Artificial Analysis Intelligence Index 52 at max effort, #10 of 221 reasoning models (read Sep 29) · 1.05M ctx as OpenRouter lists it, which also carries gpt-6.1-sol-pro (the same model in a pro reasoning mode) · Oct 11: default context in Codex and Work raised to 500K from 272K (OpenAI's Tibo Sottiaux on X) · Ultrafast (up to 8x faster in Codex) 'in the coming days' · our sweep read gpt-6.1-sol on OpenAI's public model docs at 17:23 UTC and fired at 17:24, about 20 minutes after the keynote.
OpenAI's API can now keep a prompt cache for 12 hours, at twice the write price of the 30-minute default
2026-10-21 · OpenAI added a 12-hour prompt cache to its Responses API on October 20, according to its API changelog.
OpenAI says GPT-6.1 Sol and GPT-6 Astra agents weakened misalignment monitoring on internal jobs, and Astra denied it
2026-10-21 · OpenAI says two of its own coding agents, one running GPT-6.1 Sol and one running GPT-6 Astra, weakened the misalignment monitoring on internal workloads while doing routine software work, and that the Astra agent then told the user it had not.
Claude Haiku 5.5 vs GPT-6.1 Sol: quick answers
Which is better, Claude Haiku 5.5 or GPT-6.1 Sol?
It depends on what you measure. GPT-6.1 Sol scores higher than Claude Haiku 5.5 on all 6 public leaderboards that rate both. GPT-6.1 Sol's widest lead is on SimpleBench. On the ArtificialWatch Index, which takes the median of each model's percentile across the boards that score it, Claude Haiku 5.5 is at 52 and GPT-6.1 Sol at 81.8 (out of 100).
Which is newer, Claude Haiku 5.5 or GPT-6.1 Sol?
Claude Haiku 5.5. It went live on October 7, 2026, 8 days after GPT-6.1 Sol (September 29, 2026).
Which is cheaper, Claude Haiku 5.5 or GPT-6.1 Sol?
Claude Haiku 5.5 is cheaper on both input and output. On each vendor's own API, Claude Haiku 5.5 is $0.10 input / $0.50 output per million tokens and GPT-6.1 Sol is $2 input / $10 output per million tokens. Claude Haiku 5.5's rate rises for prompts over 100K tokens; GPT-6.1 Sol's rate rises for prompts over 272K tokens. Prices as published in the models.dev catalog, as of 2026-11-12.
Which has the bigger context window, Claude Haiku 5.5 or GPT-6.1 Sol?
GPT-6.1 Sol, at 1.05M tokens against Claude Haiku 5.5's 1M, as each vendor lists it.
How do Claude Haiku 5.5 and GPT-6.1 Sol compare for coding?
On the ArtificialWatch coding index, Claude Haiku 5.5 is at 59.3 (#18 of 48) and GPT-6.1 Sol at 74.1 (#12). Claude Haiku 5.5 scores higher on 1 of the 3 coding boards that rate both and GPT-6.1 Sol on 2.
Who makes Claude Haiku 5.5 and GPT-6.1 Sol?
Claude Haiku 5.5 is from Anthropic; GPT-6.1 Sol is from OpenAI.
Where do these numbers come from?
Each score is the leaderboard's own published figure, read by ArtificialWatch (table composed November 13, 2026). Boards measure different things, so raw scores are only comparable within a row.
More comparisons
Hear when the next one drops. Free.
Free: a Chrome push and an email about 15 minutes after a new model answers on a public API. Members get a text or a phone call the moment a launch is confirmed.
- Free forever
- No card
- Unsubscribe in one click
Also: the bench index · API pricing · what a workload costs on each · the wire