Skip to main content

Model comparison

Ornith-1.0-397B vs Step 3.7 Flash

Data verified

Head-to-head evidence from 4 shared benchmark results across 2 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

DeepReinforce AI
72/100
Margin
21.2pts
← winning
50.76/100
2 category wins0 category wins

Public leaderboard positions: Ornith-1.0-397B unranked (Not scored); Step 3.7 Flash #107 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Ornith-1.0-397B and Step 3.7 Flash share 4 comparable benchmark results. 2 of 8 categories are comparable. 3 results are unique to Ornith-1.0-397B; 26 to Step 3.7 Flash.

Updated July 17, 2026
Shared results
4
Ornith-1.0-397B only
3
Step 3.7 Flash only
26
Comparable categories
2 / 8

Pick Ornith-1.0-397B if you want the stronger benchmark profile. Step 3.7 Flash only becomes the better choice if its workflow or ecosystem matters more than the raw scoreboard.

Confidence note. This is a partial-evidence comparison with 4 shared benchmark results across 2 evidence categories; 2 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Ornith-1.0-397B is clearly ahead on the BenchAlign aggregate, 72 to 50.76. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Ornith-1.0-397B's sharpest advantage is in coding, where it averages 74.6 against 56.3. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 77.5% to 59.5%.

Step 3.7 Flash is also the more expensive model on tokens at $0.20 input / $1.15 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Ornith-1.0-397B. That is roughly Infinityx on output cost alone.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for Ornith-1.0-397B and Step 3.7 Flash
CategoryOrnith-1.0-397BΔStep 3.7 Flash
CodingOrnith-1.0-397B74.6Margin 18.3Step 3.7 Flash56.3
AgenticOrnith-1.0-397B77.5Margin 11.1Step 3.7 Flash66.4

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · Ornith-1.0-397BB · Step 3.7 Flash
  1. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 77.5%B 59.5%
    Winner: Ornith-1.0-397BΔ 18
    Terminal-Bench 2.0: Ornith-1.0-397B scored 77.5%; Step 3.7 Flash scored 59.5%. Ornith-1.0-397B wins this benchmark.
  2. SWE-bench Pro

    Coding
    Source ↗
    A 62.2%B 56.3%
    Winner: Ornith-1.0-397BΔ 5.9
    SWE-bench Pro: Ornith-1.0-397B scored 62.2%; Step 3.7 Flash scored 56.3%. Ornith-1.0-397B wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricOrnith-1.0-397BStep 3.7 FlashComparison
Input / output priceUSD per 1M tokensOrnith-1.0-397B$0 input / $0 outputStep 3.7 Flash$0.2 input / $1.15 outputOrnith-1.0-397B has the lower combined listed price.
Generation speedtokens per secondOrnith-1.0-397BNot availableStep 3.7 FlashNot availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenOrnith-1.0-397BNot availableStep 3.7 FlashNot availableA complete latency comparison is not available.
Context windowmaximum listed tokensOrnith-1.0-397B256KStep 3.7 Flash256KListed context windows are equal.

Benchmark Deep Dive

AgenticOrnith-1.0-397B wins
BenchmarkOrnith-1.0-397BStep 3.7 FlashResult
Terminal-Bench 2.0Source 77.5%59.5%Ornith-1.0-397B leads
Claw-EvalSource 77.1%67.1%Ornith-1.0-397B leads
BrowseCompSource 75.8%Not comparable
DeepSearchQASource 92.8%Not comparable
GDPval-AASource 25.9%Not comparable
ToolathlonSource 49.5%Not comparable
HLE w/ toolsSource 47.2%Not comparable
Gert LabsSource 51.57%Not comparable
AA Agentic IndexSource 21.5%Not comparable
τ²-bench resultsSource 98.5%Not comparable
GDPval-AASource 1017Not comparable
APEX-Agents-AASource 14.8%Not comparable
CodingOrnith-1.0-397B wins
BenchmarkOrnith-1.0-397BStep 3.7 FlashResult
SWE-bench VerifiedSource 82.4%Not comparable
SWE-bench ProSource 62.2%56.3%Ornith-1.0-397B leads
SWE MultilingualSource 78.9%Not comparable
NL2RepoSource 48.2%Not comparable
Terminal-Bench 2.0Source 77.5%59.5%Ornith-1.0-397B leads
AA Coding IndexSource 39.6%Not comparable
Terminal-Bench HardSource 35.6%Not comparable
AA-SciCodeSource 40.0%Not comparable
Reasoning
BenchmarkOrnith-1.0-397BStep 3.7 FlashResult
AA-LCRSource 63.7%Not comparable
CritPtSource 2.3%Not comparable
Knowledge
BenchmarkOrnith-1.0-397BStep 3.7 FlashResult
Artificial Analysis Intelligence IndexSource 30.3%Not comparable
AA-GPQA DiamondSource 80.9%Not comparable
AA-HLESource 19.9%Not comparable
AA-Omniscience IndexSource -37.5%Not comparable
AA-Omniscience AccuracySource 25.4%Not comparable
AA-Omniscience Hallucination RateSource 84.4%Not comparable
Multimodal
BenchmarkOrnith-1.0-397BStep 3.7 FlashResult
SimpleVQASource 79.2%Not comparable
V*Source 95.3%Not comparable
AA-MMMU-ProSource 75.3%Not comparable
Design Arena WebsiteSource 1218Not comparable
Inst. Following
BenchmarkOrnith-1.0-397BStep 3.7 FlashResult
AA-IFBenchSource 67.3%Not comparable
Frequently Asked Questions (3)

Which is better, Ornith-1.0-397B or Step 3.7 Flash?

Ornith-1.0-397B is ahead on BenchLM's BenchAlign leaderboard, 72 to 50.76. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 77.5% and 59.5%.

Which is better for coding, Ornith-1.0-397B or Step 3.7 Flash?

Ornith-1.0-397B has the edge for coding in this comparison, averaging 74.6 versus 56.3. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Ornith-1.0-397B or Step 3.7 Flash?

Ornith-1.0-397B has the edge for agentic tasks in this comparison, averaging 77.5 versus 66.4. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 17, 2026

The AI models change fast. We track them for you.

A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.