Navigate
Advertise here

Trinity Large Thinking (medium) vs Step 5 Preview (medium)

The average score is effectively tied at 6.1 vs 6.1. Trinity Large Thinking (medium) has the lower benchmark cost at $0.862 vs $4.136. Trinity Large Thinking (medium) is faster at 84.01s vs 136.10s, with pass rates of 49.3% vs 53.6%.

Last updated at: 2026-10-09

Compared models

Rank
#246
Total Output Tokens
925,046
Response Time (avg)
84.01s
Total Cost
$0.862
Rank
#250
Total Output Tokens
1,434,621
Response Time (avg)
136.10s
Total Cost
$4.136
Recommended model Trinity Large Thinking (medium)

It has the best score here (6.1), while costing about 4.8x less than Step 5 Preview (medium).

Detailed comparison

Metric Trinity Large Thinking Trinity Large Thinking medium Release: 2026-07-28 Step 5 Preview Step 5 Preview medium Release: 2026-10-09
Score 6.1 6.1
Rank #246 #250
Reliability 10.0 9.6
Consistency 7.7 8.2
Attempts 69/69 69/69
Tests Correct
Attempt pass rate 49.3% 53.6%
Flaky tests 7 5
Total Runs 69 69
Cost per result 10.772 41.354
Total Cost $0.862 $4.136
Input Price $0.250 / 1M $1.000 / 1M
Output Price $0.800 / 1M $2.700 / 1M
Cache Read Price $0.060 / 1M $0.050 / 1M
Cache Write Price N/A N/A
Total Input Tokens 319,352 261,842
Output Tokens 158,447 1,434,621
Reasoning Tokens 875,524 0
Response Time (avg) 84.01s 136.10s
Response Time (max) 525.14s 438.09s
Response Time (total) 1932.24s 3130.27s
Parameters 398B total (13B active) 600B total (27B active)
Availability Weights available Closed

Cache prices apply to input tokens. Reads reuse cached prompts; writes store them and can cost extra. Output tokens use the output price.

Model generation showcase

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#246 Trinity Large Thinking

medium
Cost
$0.016
Time
75.9s
Tokens
19,401 tok

#250 Step 5 Preview

medium
Cost
$0.042
Time
90.3s
Tokens
15,574 tok

Top Models by Score

Score vs Total Cost

Response Time (avg)

Score vs Response Time (avg)

Total Output Tokens

Score vs Total Output Tokens

Category Breakdown

Coding Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Input Tokens Output Tokens Reasoning Tokens
Trinity Large Thinking 7.5 10.0 66.7% 0 179.02s 7,248 7,413 351,155
Step 5 Preview 2.8 7.2 11.1% 1 363.90s 7,428 546,725 0

Quick Compare

Switch Comparison Pair

Qwen3.5 Plus 2026-04-20nonevsStep 5 PreviewmediumStep 5 PreviewmediumvsInklingnoneFree AvailableApodex 1.1 MinilowFree AvailablevsStep 5 PreviewmediumApodex 1.1 MinilowFree AvailablevsTrinity Large ThinkingmediumTrinity Large ThinkingmediumvsInklingnoneFree AvailableTrinity Large ThinkingmediumvsQwen3.5 Plus 2026-04-20noneKAT-Coder-Pro V2.5defaultvsStep 5 PreviewmediumTrinity Large ThinkingmediumvsGemini 3.5 FlashnoneTrinity Large ThinkingmediumvsKAT-Coder-Pro V2.5defaultGemini 3.5 FlashnonevsStep 5 PreviewmediumClaude Fable 5.1lowvsTrinity Large ThinkingmediumTrinity Large ThinkingmediumvsGemini 3.1 Flash Lite Previewnone