Claude Haiku 5.5 vs GPT-6.1 Terra

aka Haiku 5.5 vs OpenAI GPT-6.1 Terra

7 posts151 likes14.1K views

  • CheaperGPT-6.1 Terra2/2 postsmedian gap ~2.5×
  • FasterGPT-6.1 Terra3/3 postsmedian gap ~4.5×

As stated in each post, not a controlled benchmark.

Posts

7

Same prompt, separate runs

5

Not a controlled comparison: prompts, tools and time differ.

About this matchup

Claude Haiku 5.5 vs GPT-6.1 Terra in real side-by-side tests: 7 posts from X where both got the same prompt; 3 posts report cost or time. Judge coding ability from the actual outputs.

Haiku 5.5 vs GPT-6.1 Terra: which is better for coding?

Which is better for coding, Haiku 5.5 or GPT-6.1 Terra?

None of the 7 posts here names one of the two as the better result, so judge the outputs above for the work you care about.

Which is cheaper, Haiku 5.5 or GPT-6.1 Terra?

2 posts state a cost for both: GPT-6.1 Terra was cheaper in both (median gap ~2.5×). These are the costs the posts state for each run, not list prices.

Which is faster, Haiku 5.5 or GPT-6.1 Terra?

3 posts state a time for both: GPT-6.1 Terra was faster in all 3 (median gap ~4.5×). Times are as stated in each post; tools and settings differ between posts.

Summary

Claude Haiku 5.5 and GPT-6.1 Terra appear together in 7 posts on X: rebuild a 75 s Short in fframes and Remotion (@techfren), vibe video editing, four Claude and GPT-6.1 models (@nfcodes), FawnBench running deer, four budget models (@fawnunwrap), campfire, GPT-6.1 Terra vs Claude Haiku 5.5 (@rubenssoto_ai), Three.js galloping zebra, three models (@AI_Screening) and 2 more. 3 of them are part of larger lineups with other models. In @MrTacticalX's snow globe winter village in three.js, Terra vs Haiku, GPT-6.1 Terra was cheaper ($0.13 vs $0.32) and faster (2 min vs 9 min), as reported in the post. In @aGim_asf's Minecraft clone in one HTML file, Terra vs Haiku, GPT-6.1 Terra was faster (10 min vs 29 min), as reported in the post. In @fawnunwrap's FawnBench running deer, four budget models, GPT-6.1 Terra was cheaper ($0.032 vs $0.081), faster (35 s vs 10.3 min) and used fewer tokens (21K tokens vs 205K tokens), as reported in the post.

What each post reports

Only numbers stated in each post. “—” means the post doesn’t say.

Quick facts

ModelVendorReleasedInput $/1M tokensOutput $/1M tokensContext (tokens)Official links
Claude Haiku 5.5Anthropic
GPT-6.1 TerraOpenAI2026-10-26Announcement

From the vendors’ official pages, checked 2026-10-26. Standard API list prices; each value links its source and a blank cell could not be verified. The archive’s own numbers above are what each post states.

Who won where

1

Models

Related matchups