Skip to main content

AnthropicMoonshot AILLM matchup

Claude Haiku 5.5vs.Kimi K3

Which one is better?

Claude Haiku 5.5 and Kimi K3 perform about equally well, but Kimi K3 costs 30 times as much.

level
IntelligenceOverall index, 100 points = average
103.4

Claude Haiku 5.5

Rank 20

101.2

Kimi K3

Rank 23

Haiku 5.5 faster
SpeedTokens per second
236

Claude Haiku 5.5

39

Kimi K3

Haiku 5.5 cheaper
Price$ per 1M tokens
$0.20

Claude Haiku 5.5

$6

Kimi K3

Haiku 5.5 more popular
Popularity0 to 100
77

Claude Haiku 5.5

Rank 3 of 875

68

Kimi K3

Rank 19 of 875

Above 100,000 tokens of context, Claude Haiku 5.5 costs more: $0.50 input and $2.50 output per 1M tokens.

Choose Claude Haiku 5.5 for

CodingAgentsEducationMathematicsHealthReasoning (not certain)

Choose Kimi K3 for

ProductivityLegal (not certain)Knowledge (not certain)

Where is the difference?

Strengths at a glance

Haiku 5.5Kimi K3

The median of all ranked models is below 100 in every area, so it sits in the centre of the net.

105110100CodingHaiku 5.5Reasoningslightly Haiku 5.5MathematicsHaiku 5.5Knowledgeslightly Kimi K3AgentsHaiku 5.5GenerallevelFinancelevelLegalslightly Kimi K3HealthHaiku 5.5
CodingHaiku 5.5 ahead
Haiku 5.5106.2Rank 18
Kimi K397.3Rank 37
Median of ranked models83.7

95 % range: Haiku 5.5 102.5 to 109.9, Kimi K3 90.7 to 103.9.

AreaHaiku 5.5Kimi K3

The same values and verdicts as the bars below. In the top field the net starts at 100 points, the reference group average; values below sit as a hollow point in the centre. Not in the net because a model has no own value there or the data are too thin: Multimodal.

Haiku 5.5 is ahead in 4 of 10 areas, Kimi K3 in none.

Haiku 5.5 ahead: 4Haiku 5.5 slightly ahead: 1level: 2too little data: 1Kimi K3 slightly ahead: 2

Areas in the overall index
CodingHaiku 5.5 +9106.2 to 97.3
AgentsHaiku 5.5 +9106.1 to 97.4
MathematicsHaiku 5.5 +5107.1 to 101.7
Knowledgeleaning Kimi K3not certain · 2 and 3 tests
Reasoningleaning Haiku 5.5not certain · 4 and 5 tests
Generallevel106.4 to 104.4
Multimodalonly Kimi K3 has a score
Specialist fields
HealthHaiku 5.5 +5108.9 to 103.9
Legalleaning Kimi K3not certain · 3 and 4 tests
Financelevel104.7 to 104.9

Bar length shows the gap in index points (full width 15 points, one standard deviation of the reference group). Solid means clearly ahead, hatched not certain yet, a thin grey bar level.

Show 3 more areas that do not count toward the overall index
EducationHaiku 5.5 +9109.6 to 101.1
ProductivityKimi K3 +3101.8 to 105.1
Cybersecuritytoo little data108.6 to 105.0

Bar length shows the gap in index points (full width 15 points, one standard deviation of the reference group). Solid means clearly ahead, hatched not certain yet, a thin grey bar level.

Claude Haiku 5.5 vs. Kimi K3 at the same effort

Both models measured at the same reasoning efforts. This shows whether a lead comes from the model or only from more compute.

lowtied
Haiku 5.5 80.8 vs. Kimi K3 86.9 · Gap 1.1 · Test series 11
maxHaiku 5.5 slightly ahead
Haiku 5.5 103.4 vs. Kimi K3 101.2 · Gap 6.4 · Test series 3
Overall index (max vs. max)level
Haiku 5.5 103.4 vs. Kimi K3 101.2

Gap and verdict compare only the Artificial Analysis test series in which both models are measured at that effort. Tied below 3 points, ahead only beyond two standard errors. The single values rest on all test series of each model, so their difference can deviate from the gap.

Which model for which task?

Everyday tasks, translated from the area values above. Each row names the values it rests on.

Writing and debugging codeHaiku 5.5Coding 106.2 to 97.3
Maths and data analysisHaiku 5.5Mathematics 107.1 to 101.7, Reasoning 105.2 to 97.5
Facts and researchleaning Kimi K3Knowledge 85.5 to 95.9
Agents and workflowsHaiku 5.5Agents 106.1 to 97.4
Fast chat repliesHaiku 5.5236 to 39 tokens per second
Security analysisopenCybersecurity 108.6 to 105.0 (too little data)
Many requests on a small budgetHaiku 5.5$0.20 to $6 per 1M tokens
Finance, legal and medical textsdepends on the partFinance 104.7 to 104.9, Legal 94.4 to 108.5, Health 108.9 to 103.9
Images, screenshots and PDFsopenMultimodal without a score for Haiku 5.5

See for yourself: same task, one attempt

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

Prompt

Return only one standalone SVG of a pelican riding a bicycle. Use viewBox="0 0 800 600", simple shapes, two clearly visible wheels and a recognizable pelican with a long beak. At most 30 SVG elements and 1800 characters. No Markdown, explanations, comments, scripts, animation, CSS or external resources.

Claude Haiku 5.5

Unavailable in this round

Kimi K3

Kimi K3: SVG short test

Oct 14, 2026 · $0.0049

Test details
Test run
Oct 14, 2026
Provider
chutes/mxfp4
API model
moonshotai/kimi-k3
Output
495 tokens
Reasoning
API reports 0 reasoning tokens
Cost of the attempt
$0.0049
Test round
43b8c5c46830
Earlier tests

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Task origin: Simon Willison

Claude Haiku 5.5

Not tested yet

Kimi K3

Kimi K3: Pelican on a bicycle
A community classic for free-form SVG drawing.

Oct 1, 2026 · 173 s · $0.08

Test details
Run date
Oct 1, 2026
API provider
openrouter
Requested API model
moonshotai/kimi-k3
Reasoning setting
high
Token limit including reasoning
8,192 tokens
Output tokens
5,145 tokens
Cost of the attempt
$0.0775
Response time
173.4 s
API model reported in response
moonshotai/kimi-k3
Pinned endpoint
Moonshot AI | moonshotai/kimi-k3-20260715
Temperature
Provider default
Test protocol
community-visual-tests-v1
Technical checks
Passed, not a quality rating

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (1)

  • Sep 7, 2026API request failedReasoning setting: high

Claude Haiku 5.5 and Kimi K3 in the overall index

Every model with a place in the overall index, both models of this matchup highlighted. You can add or hide other models.

Overall index (Gradually Meta-Score v4)

20 of 91 models

4 models in this view are measured at a lower reasoning effort than max in the overall index. Dashed column: lower reasoning effort than the top.

Claude Opus 5.5max126.1
Gemini 4 Carbonhigh125.7
Claude Fable 5.1max123.8
Claude Sonnet 5.5max122.8
Gemini 4 Argonhigh122.3
GPT-6 Astramax121.3
GPT-6.1 Solmax119.2
Muse Wonder Fullmax118.3
Claude Opus 5max117.2
Muse Wondermax113.4
Kimi K3.1 Promax112.9
Claude Fable 5max112.5
GPT-6 Solmax110.6
GPT-5.6 Solmax108.7
Muse Spark 1.3max107.4
Gemini 3.7 Flashhigh107.1
DeepSeek V4.1 Flash 1017max106.6
Gemini 3.8 Flashhigh106.5
Claude Haiku 5.5max103.4
Kimi K3max101.2

Popularity compared

How often the models are searched for, used and mentioned, on a scale from 0 to 100.

Claude Haiku 5.5 is more popular than Kimi K3: 77 to 68 out of 100.

Over time

Weekly values from October 18, 2026 to November 5, 2026

Claude Haiku 5.5Kimi K3

All data side by side

The better value is bold. For the overall index only when the gap is clear.

MetricClaude Haiku 5.5Kimi K3
Intelligence
Overall indexpoints, 100 = reference group average, higher is better103.4101.2
RankRank 20 of 91Rank 23 of 91
Price per 1M tokens
Inputwhat you send$0.10above 100,000 tokens: $0.50$3
Outputwhat the model writes$0.50above 100,000 tokens: $2.50$15
Speed
Tokens per second23639
Popularity
Popularitysearch, usage and mentions, 0 to 10077Rank 3 of 87568Rank 19 of 875
Technical
Context windowhow much text the model reads at once1 M tokens≈ 1,500 A4 pages1.05 M tokens≈ 1,570 A4 pages
Knowledge cutoffJune 2026n/a
ReleasedOct 7, 2026Jul 16, 2026
AvailabilityActiveActive
Open weightsnoyes

Prices without caching or batch discounts · Page count as a rule of thumb from Artificial Analysis

Is there a better choice?

Successor of Kimi K3Kimi K3.1 Pro

Overall index 112.9, rank 11. $2 input and $10 output per 1M tokens.

Haiku 5.5 vs. Kimi K3.1 Pro
Best model in the overall indexClaude Opus 5.5

Overall index 126.1, rank 1. $4 input and $20 output per 1M tokens.

Haiku 5.5 vs. Claude Opus 5.5

Shared benchmarks

Single results from the same benchmark version, task and metric. A value is bold only with the same test setup and a clear gap.

A: Claude Haiku 5.5B: Kimi K3

AA-AnalystAgentOverall · higher is betterClaude Haiku 5.5: 53.8%Kimi K3: 38.8%Artificial AnalysisAs of 11/06/2026
AA-BriefcaseOverall · higher is betterClaude Haiku 5.5: 1,114.41 eloKimi K3: 985.1 eloArtificial AnalysisAs of 11/06/2026
AA-OmniscienceOverall · higher is betterClaude Haiku 5.5: 2.167 pointsKimi K3: -1.7 pointsArtificial AnalysisAs of 11/06/2026
AA-Omniscience AccuracyOverall · higher is betterClaude Haiku 5.5: 30.4%Kimi K3: 42.4%Artificial AnalysisAs of 11/06/2026
AA-Omniscience Hallucination RateOverall · lower is betterClaude Haiku 5.5: 40.5%Kimi K3: 76.6%Artificial AnalysisAs of 11/06/2026
AutomationBench-AAOverall · higher is betterClaude Haiku 5.5: 22.9%Kimi K3: 49.3%Artificial AnalysisAs of 11/06/2026
BioMysteryBenchOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 70.4%Kimi K3: 72.6%Vals AIAs of 11/05/2026
Code MigrationOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 44.6%Kimi K3: 16.1%Vals AIAs of 10/29/2026
Code Migration v1.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 52.4%Kimi K3: 28.6%Vals AIAs of 10/30/2026
CritPt v2 (Hard Subset)Overall · higher is betterClaude Haiku 5.5: 21.5%Kimi K3: 5.5%Artificial AnalysisAs of 11/06/2026
Show 54 more benchmarks
CyberBench v1.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 75.5%Kimi K3: 75.2%Vals AIAs of 11/05/2026
EMBOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 59%Kimi K3: 66.7%Vals AIAs of 10/20/2026
EMB v1.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 65.5%Kimi K3: 71.3%Vals AIAs of 11/05/2026
Finance Agent (v2)Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 54.1%Kimi K3: 53.1%Vals AIAs of 10/29/2026
Finance Agent v2.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 65.4%Kimi K3: 60%Vals AIAs of 11/05/2026
FrontierCodeOverall · higher is betterClaude Haiku 5.5: 8.0%Kimi K3: 10.9%Artificial AnalysisAs of 11/06/2026
FrontierCode 1.1 MainOverall · higher is betterClaude Haiku 5.5: 29.3%Kimi K3: 28.5%Artificial AnalysisAs of 11/06/2026
GDP.pdfOverall · higher is betterClaude Haiku 5.5: 11.2%Kimi K3: 18.6%Artificial AnalysisAs of 11/06/2026
GDPval-AAOverall · higher is betterClaude Haiku 5.5: 1,130.36 eloKimi K3: 1,135.77 eloArtificial AnalysisAs of 11/06/2026
HLAB-Vals-1.1Overall · Task fully resolved · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 4.2%Kimi K3: 12.9%Vals AIAs of 11/05/2026
HLE-DiamondOverall · higher is betterClaude Haiku 5.5: 14.1%Kimi K3: 14.0%Artificial AnalysisAs of 11/06/2026
ImmigrationOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 61.8%Kimi K3: 68.3%Vals AIAs of 10/28/2026
IOIOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 47.3%Kimi K3: 48.9%Vals AIAs of 10/29/2026
Legal Research BenchOverall · All-pass · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 43.8%Kimi K3: 46.2%Vals AIAs of 11/05/2026
LMArena Code Arena WebDevOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 1,589Kimi K3: 1,656LMArenaAs of 10/28/2026
LMArena Text, Style ControlOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 1,425Kimi K3: 1,488LMArenaAs of 10/28/2026
MedCodeOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 51.5%Kimi K3: 49.4%Vals AIAs of 10/29/2026
MedScribeOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 89.5%Kimi K3: 88%Vals AIAs of 10/29/2026
ProgramBenchOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 2.5%Kimi K3: 2%Vals AIAs of 11/05/2026
ProofBench v1.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 89%Kimi K3: 87%Vals AIAs of 10/29/2026
Public Benefits Bench v1.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 63.2%Kimi K3: 68.2%Vals AIAs of 10/12/2026
SAGEOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 89.8%Kimi K3: 86.8%Vals AIAs of 11/05/2026
SWE-bench MultimodalOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 44%Kimi K3: 44.2%Vals AIAs of 11/05/2026
Tax Agent BenchOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 31.9%Kimi K3: 33.4%Vals AIAs of 11/05/2026
Terminal-BenchOverall · higher is betterClaude Haiku 5.5: 12.6%Kimi K3: 12.6%Artificial AnalysisAs of 11/06/2026
Terminal-Bench 4.0Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 36.4%Kimi K3: 17.2%Vals AIAs of 11/05/2026
Terminal-Bench-ScienceOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 10%Kimi K3: 1.4%Vals AIAs of 11/05/2026
Terminal-Bench-ScienceOverall · higher is betterClaude Haiku 5.5: 1.9%Kimi K3: 1.0%Artificial AnalysisAs of 11/06/2026
Vals Finance IndexOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 57.4%Kimi K3: 58%Vals AIAs of 11/05/2026
Vals IndexOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 58.7%Kimi K3: 53%Vals AIAs of 11/05/2026
Vals Legal IndexOverall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 70.3%Kimi K3: 71.7%Vals AIAs of 11/05/2026
Vals Router IndexFull Index · higher is betterClaude Haiku 5.5: 58.7%Kimi K3: 53%Vals AIAs of 11/03/2026
Vibe Code Bench 1-100Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 20.4%Kimi K3: 18.2%Vals AIAs of 10/29/2026
Vibe Code Bench v1.1Overall · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 90.4%Kimi K3: 85%Vals AIAs of 11/05/2026
BioMysteryBenchHard · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 29.4%Kimi K3: 41.2%Vals AIAs of 11/05/2026
Code MigrationCLI · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 36.1%Kimi K3: 1.5%Vals AIAs of 10/29/2026
Code Migration v1.1CLI · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 42.1%Kimi K3: 11.4%Vals AIAs of 10/30/2026
CyberBench v1.1Patch · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 89.3%Kimi K3: 82.1%Vals AIAs of 11/05/2026
EMBComps · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 77.9%Kimi K3: 70.8%Vals AIAs of 10/20/2026
EMB v1.1Comps · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 77.7%Kimi K3: 76.4%Vals AIAs of 11/05/2026
Finance Agent (v2)Adjustments · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 49.6%Kimi K3: 45.3%Vals AIAs of 10/29/2026
Finance Agent v2.1Adjustments · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 67.7%Kimi K3: 61.8%Vals AIAs of 11/05/2026
FrontierCode 1.1 MainOverall · higher is betterClaude Haiku 5.5: 41.8%Kimi K3: 43.6%Artificial AnalysisAs of 11/06/2026
GDPval-AAOverall · higher is betterClaude Haiku 5.5: 1,630.83 eloKimi K3: 1,542.76 eloArtificial AnalysisAs of 11/06/2026
HLAB-Vals-1.1Antitrust Competition · Criteria passed · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 86.2%Kimi K3: 93.1%Vals AIAs of 11/05/2026
ImmigrationAll-Pass · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 16%Kimi K3: 21.4%Vals AIAs of 10/28/2026
IOIIOI 2024 · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 67.7%Kimi K3: 59.7%Vals AIAs of 10/29/2026
Legal Research BenchAdministrative / Regulatory · All-pass · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 50%Kimi K3: 63.6%Vals AIAs of 11/05/2026
LMArena Code Arena WebDevHTML development · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 1,568Kimi K3: 1,641LMArenaAs of 10/28/2026
LMArena Text, Style ControlBusiness, management and finance · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 1,435Kimi K3: 1,487LMArenaAs of 10/28/2026
ProgramBenchAlmost Resolved · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 29.5%Kimi K3: 17.5%Vals AIAs of 11/05/2026
SAGEComputer Science · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 91.6%Kimi K3: 91.6%Vals AIAs of 11/05/2026
SWE-bench Multimodalbpmn-js · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 64.2%Kimi K3: 56.6%Vals AIAs of 11/05/2026
Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identicalClaude Haiku 5.5: 6.7%Kimi K3: 13.3%Vals AIAs of 11/05/2026

More comparisons

Compare Claude Haiku 5.5 and Kimi K3 with the models people compare them with most often.

Compare another matchup

The most popular comparisons

  1. 1Claude Haiku 5.5 vs. GPT-6 Luna
  2. 2Claude Sonnet 5.5 vs. Claude Haiku 5.5
  3. 3Claude Haiku 5.5 vs. DeepSeek V4.1 Flash 1017
  4. 4Claude Haiku 5.5 vs. DeepSeek-V4-Flash
  5. 5Claude Opus 5.5 vs. Kimi K3
  6. 6Claude Haiku 5.5 vs. DeepSeek-V4.1-Flash
  7. 7GPT-6 Astra vs. Kimi K3
  8. 8Claude Opus 5.5 vs. Claude Haiku 5.5
  9. 9Claude Haiku 5.5 vs. GLM-5.3
Show all 52 comparisons
  1. 10Claude Haiku 5.5 vs. Gemini 3.5 Flash-Lite
  2. 11Claude Haiku 5.5 vs. Gemini 3.8 Flash
  3. 12Claude Sonnet 5.5 vs. Kimi K3
  4. 13Claude Haiku 5.5 vs. Claude Sonnet 4.5
  5. 14Grok 4.7 vs. Kimi K3
  6. 15Gemini 3.1 Pro Preview vs. Kimi K3
  7. 16Claude Haiku 5.5 vs. Claude Sonnet 5
  8. 17Claude Haiku 5.5 vs. Claude Opus 5
  9. 18Claude Haiku 5.5 vs. GPT-6 Astra
  10. 19Claude Haiku 5.5 vs. Gemini 3.5 Flash
  11. 20Claude Opus 5 vs. Kimi K3
  12. 21Claude Haiku 5.5 vs. Claude Opus 4.8
  13. 22Kimi K3 vs. GLM-5.3-Flash
  14. 23GPT-6.1 Sol vs. Kimi K3
  15. 24Claude Haiku 5.5 vs. Gemini 4 Argon
  16. 25Claude Haiku 5.5 vs. Grok 4.7
  17. 26DeepSeek-V4-Pro vs. Kimi K3
  18. 27Claude Haiku 5.5 vs. GPT-5.6 Luna
  19. 28Claude Haiku 5.5 vs. Qwen 3.8 Max 0902
  20. 29Claude Haiku 5.5 vs. Muse Spark 1.3
  21. 30Claude Haiku 5.5 vs. Claude Sonnet 4.6
  22. 31Claude Haiku 5.5 vs. Gemini 3.7 Flash
  23. 32Claude Haiku 5.5 vs. Mistral Large 4
  24. 33Kimi K3 vs. GLM-5.3
  25. 34Claude Haiku 5.5 vs. GPT-6.1 Sol
  26. 35GPT-6 Sol vs. Kimi K3
  27. 36Claude Haiku 5.5 vs. Gemini 3.1 Pro Preview
  28. 37Claude Haiku 5.5 vs. GLM-5.3-Flash
  29. 38Claude Haiku 5.5 vs. DeepSeek-V4-Pro
  30. 39Claude Haiku 5.5 vs. GPT-5.6 Sol
  31. 40Claude Haiku 5.5 vs. Gemini 3.6 Flash
  32. 41DeepSeek-V4.1-Flash vs. Kimi K3
  33. 42Claude Haiku 5.5 vs. Claude Opus 4.6
  34. 43Claude Haiku 5.5 vs. Kimi K2.6
  35. 44Claude Fable 5 vs. Kimi K3
  36. 45Kimi K3 vs. Mistral Large 4
  37. 46Claude Opus 4.8 vs. Kimi K3
  38. 47Claude Haiku 5.5 vs. Claude Opus 4.7
  39. 48GPT-5.4 vs. Kimi K3
  40. 49Claude Haiku 5.5 vs. Mistral Large 3
  41. 50Claude Haiku 5.5 vs. GPT-5.5
  42. 51Muse Wonder Full vs. Kimi K3
  43. 52Claude Haiku 5.5 vs. GPT-6.1 Terra