Model vs model

Step 3.5 Flash vs Agnes 2.5 Pro Alpha

A debate-ready pair page: current winner, strongest alternative, decisive benchmarks, and the warning that should travel with the claim.

Use case · Everyday chatbot
Winner · Step 3.5 Flash
Sources · All public sources

Winner

Step 3.5 Flash

StepFun

0benchmarks won

Search / tool use
Chat / text

Versus · Everyday chatbot

Step 3.5 Flash leads this compare set for everyday chatbot.

Step 3.5 Flash0 of 0 wins

Thin verified coverage

0 shared benchmarks are still too close to call, so the win stays conditional. This compare uses all public sources, with provider-official evidence labeled separately.

Close calls: 0
Sources: All public

Full compare workspace Disagreement page

0%win share

Challenger

Agnes 2.5 Pro Alpha

Unknown

0benchmarks won

Reasoning / math / science

The cases in full

Step 3.5 Flash case

Search / tool use
Chat / text

Agnes 2.5 Pro Alpha case

Reasoning / math / science

What changes the outcome

Step 3.5 Flash: 120 visible benchmark gaps still leave room for the result to move.
Agnes 2.5 Pro Alpha: 121 visible benchmark gaps still leave room for the result to move.

Why this result is surprising

The visible shared evidence is more decisive than usual for this compare set.
Very few shared benchmarks are decisively separating these models.

Why this is not a clean win

0 shared benchmarks are still too close to call, so the win stays conditional. This compare uses all public sources, with provider-official evidence labeled separately.
Agnes 2.5 Pro Alpha remains the strongest alternative once you change use case, mode, or missing-evidence assumptions.

Open full compare workspace Open evidence page Open disagreement page

Advanced framings and X composerNeutral, contrarian, open-model, and skeptical variants

Neutral analystLead with the recommendation, then attach the reason and warning.Step 3.5 Flash leads this compare set for everyday chatbot.

ContrarianPush against the easy read and keep the strongest alternative live.Contrarian take: Step 3.5 Flash leads this compare set for everyday chatbot.

Open-model angleBias the framing toward open weights and source transparency.Open-model angle: Model compare · Step 3.5 Flash vs Agnes 2.5 Pro Alpha

Don't trust the headlineLead with the warning before you let the claim travel.Don't trust the headline: Model compare · Step 3.5 Flash vs Agnes 2.5 Pro Alpha

Decisive benchmarks

14 of 138 benchmarks


Intelligence Index AA · index Text · Chat / text	26n/a exact aliasverified runtimeContext only Row details Raw value 26 Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	40n/a exact aliasverified runtimeContext only Row details Raw value 40 Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Time to first token AA · s Text · Chat / text	1.12sn/a exact aliasverified runtimeContext only Row details Raw value 1.12s Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	2.09sn/a exact aliasverified runtimeContext only Row details Raw value 2.09s Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Output Speed AA · tokens/s Text · Chat / text	218.3 tokens/sn/a exact aliasverified runtimeContext only Row details Raw value 218.3 tokens/s Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	132 tokens/sn/a exact aliasverified runtimeContext only Row details Raw value 132 tokens/s Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Time to first answer token AA · s Text · Chat / text	10.28sn/a exact aliasverified runtimeContext only Row details Raw value 10.28s Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	17.24sn/a exact aliasverified runtimeContext only Row details Raw value 17.24s Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Blended price AA · /1M tokens Text · Chat / text	$0.2 /1M tokensn/a exact aliasverified runtimeContext only Row details Raw value $0.2 /1M tokens Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	$0.6 /1M tokensn/a exact aliasverified runtimeContext only Row details Raw value $0.6 /1M tokens Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Input price AA · /1M input tokens Text · Chat / text	$0.1 /1M input tokensn/a exact aliasverified runtimeContext only Row details Raw value $0.1 /1M input tokens Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	$0.5 /1M input tokensn/a exact aliasverified runtimeContext only Row details Raw value $0.5 /1M input tokens Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Output price AA · /1M output tokens Text · Chat / text	$0.3 /1M output tokensn/a exact aliasverified runtimeContext only Row details Raw value $0.3 /1M output tokens Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	$0.9 /1M output tokensn/a exact aliasverified runtimeContext only Row details Raw value $0.9 /1M output tokens Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
GPQA AA · % Text · Reasoning / math / science	83.1%n/a exact aliasverified runtimeContext only Row details Raw value 83.1% Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	87.6%n/a exact aliasverified runtimeContext only Row details Raw value 87.6% Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
Humanity's Last Exam AA · % Text · Reasoning / math / science	21.1%n/a exact aliasverified runtimeContext only Row details Raw value 21.1% Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	33.6%n/a exact aliasverified runtimeContext only Row details Raw value 33.6% Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
CritPt AA · % Text · Reasoning / math / science	2.5%n/a exact aliasverified runtimeContext only Row details Raw value 2.5% Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	10.9%n/a exact aliasverified runtimeContext only Row details Raw value 10.9% Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
SciCode AA · % Code · Coding	40.4%n/a exact aliasverified runtimeContext only Row details Raw value 40.4% Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	42.2%n/a exact aliasverified runtimeContext only Row details Raw value 42.2% Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a
AA-Omniscience accuracy AA · % Text · Chat / text	23.6%n/a exact aliasverified runtimeContext only Row details Raw value 23.6% Percentile n/a Last updated recent Eligibility benchmark_derived_model Identity provider alias (0.94) Source label Step 3.5 Flash Source row	33.5%n/a exact aliasverified runtimeContext only Row details Raw value 33.5% Percentile n/a Last updated recent Eligibility preview_model Identity provider alias (0.94) Source label Agnes 2.5 Pro Alpha Source row	n/a

1-12 of 14

Page 1 of 2Page size

Intelligence Index

AA · index

Text · Chat / text

26n/a

exact aliasverified runtimeContext only

Row details

Raw value: 26
Percentile: n/a
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.94)
Source label: Step 3.5 Flash

Step 3.5 Flash vs Agnes 2.5 Pro Alpha

Step 3.5 Flash

Step 3.5 Flash leads this compare set for everyday chatbot.

Agnes 2.5 Pro Alpha

Step 3.5 Flash case

Agnes 2.5 Pro Alpha case

What changes the outcome

Why this result is surprising

Why this is not a clean win

Share the recommendation with the source data attached.

Open or copy the stable surfaces

Use the exact public framing

Pick the voice before you post

Compose a post that keeps the warning attached

Decisive benchmarks

Loading model comparison.

Step 3.5 Flash vs Agnes 2.5 Pro Alpha

Step 3.5 Flash

Step 3.5 Flash leads this compare set for everyday chatbot.

Agnes 2.5 Pro Alpha

Step 3.5 Flash case

Agnes 2.5 Pro Alpha case

What changes the outcome

Why this result is surprising

Why this is not a clean win

Share the recommendation with the source data attached.

Open or copy the stable surfaces

Use the exact public framing

Pick the voice before you post

Compose a post that keeps the warning attached

Decisive benchmarks