SUPERChem — Mercor evaluation (SUPERChem (Mercor))

We show this table for reference; we do not rank on it.

Answering advanced chemistry questions. This table shows Mercor-run configurations for reference and is excluded from model rankings.

Evaluation results published by Mercor. Published evaluation aggregates only; task contents and dataset license grants are not included.

Pass@1 on SUPERChem (Mercor) — October 15, 2026 capture

We mirror the published pass@1 view for SUPERChem (Mercor). GPT-6 Astra leads the public snapshot at 80.30%, followed by Opus 5.5 (79.60%) and GPT-6.1 Sol (77.80%). We do not use these results to rank models overall.

GPT-6 Astra

OpenAI

Single-shot multimodal · xhigh reasoning

Published margin ±3.2 · 1,500 reported samples

80.30%

Opus 5.5

Anthropic

Single-shot multimodal · max reasoning

Published margin ±3.3 · 1,500 reported samples

79.60%

GPT-6.1 Sol

OpenAI

Single-shot multimodal · max reasoning

Published margin ±3.3 · 1,472 reported samples

77.80%

33 configurationsKnowledgeDated Mercor configurationsDisplay onlyUpdated October 15, 2026 capture

Pass@1 table (33 configurations)

Pass@1 results for SUPERChem (Mercor)
Rank	Model / configuration	Score	Parameters (B)	Open / closed
1	GPT-6 AstraOpenAISingle-shot multimodal · xhigh reasoningPublished margin ±3.2 · 1,500 reported samples	80.30%	Not reported	Unknown
2	Opus 5.5AnthropicSingle-shot multimodal · max reasoningPublished margin ±3.3 · 1,500 reported samples	79.60%	Not reported	Unknown
3	GPT-6.1 SolOpenAISingle-shot multimodal · max reasoningPublished margin ±3.3 · 1,472 reported samples	77.80%	Not reported	Unknown
4	GPT-6 AstraOpenAISingle-shot multimodal · max reasoningPublished margin ±3.4 · 1,500 reported samples	75.90%	Not reported	Unknown
5	Sonnet 5.5AnthropicSingle-shot multimodal · max reasoningPublished margin ±3.6 · 1,403 reported samples	75.40%	Not reported	Unknown
6	GPT-5.6 Sol (Pro)OpenAISingle-shot multimodal · max reasoningPublished margin ±3.5 · 1,495 reported samples	74.90%	Not reported	Unknown
7	Opus 5AnthropicSingle-shot multimodal · max reasoningPublished margin ±3.5 · 1,500 reported samples	74.70%	Not reported	Unknown
8	Fable 5.1AnthropicSingle-shot multimodal · max reasoningPublished margin ±3.5 · 1,500 reported samples	74.70%	Not reported	Unknown
9	Gemini 3.7 FlashGoogleSingle-shot multimodal · high reasoningPublished margin ±3.6 · 1,500 reported samples	72.80%	Not reported	Unknown
10	GPT-6 SolOpenAISingle-shot multimodal · max reasoningPublished margin ±3.6 · 1,500 reported samples	72.80%	Not reported	Unknown
11	Gemini 3.8 FlashGoogleSingle-shot multimodal · high reasoningPublished margin ±3.5 · 1,499 reported samples	72.60%	Not reported	Unknown
12	GPT-5.6 SolOpenAISingle-shot multimodal · xhigh reasoningPublished margin ±3.7 · 1,500 reported samples	70.90%	Not reported	Unknown
13	Gemini 3.5 FlashGoogleSingle-shot multimodal · high reasoningPublished margin ±3.6 · 1,500 reported samples	68.30%	Not reported	Unknown
14	GPT-5.5OpenAISingle-shot multimodal · xhigh reasoningPublished margin ±3.8 · 1,487 reported samples	66.80%	Not reported	Unknown
15	Gemini 3.6 FlashGoogleSingle-shot multimodal · high reasoningPublished margin ±3.7 · 1,497 reported samples	66.00%	Not reported	Unknown
16	Grok 4.6xAISingle-shot multimodal · xhigh reasoningPublished margin ±3.8 · 1,500 reported samples	65.70%	Not reported	Unknown
17	Grok 4.5xAISingle-shot multimodal · high reasoningPublished margin ±3.8 · 1,500 reported samples	65.50%	Not reported	Unknown
18	GPT-5.6 TerraOpenAISingle-shot multimodal · max reasoningPublished margin ±3.7 · 1,500 reported samples	64.90%	Not reported	Unknown
19	GPT 5.4OpenAISingle-shot multimodal · high reasoningPublished margin ±3.7 · 1,500 reported samples	64.10%	Not reported	Unknown
20	Kimi K3KimiSingle-shot multimodal · max reasoningPublished margin ±3.7 · 1,500 reported samples	63.60%	Not reported	Unknown
21	Fable 5AnthropicSingle-shot multimodal · max reasoningPublished margin ±4 · 1,500 reported samples	62.80%	Not reported	Unknown
22	Qwen3.8-MaxAlibabaSingle-shot multimodal · xhigh reasoningPublished margin ±3.8 · 1,499 reported samples	62.30%	Not reported	Unknown
23	Haiku 5.5AnthropicSingle-shot multimodal · max reasoningPublished margin ±3.7 · 1,500 reported samples	62.00%	Not reported	Unknown
24	Opus 4.8AnthropicSingle-shot multimodal · max reasoningPublished margin ±4 · 1,500 reported samples	60.30%	Not reported	Unknown
25	Fable 5.1AnthropicSingle-shot multimodal · high reasoningPublished margin ±3.9 · 1,500 reported samples	59.30%	Not reported	Unknown
26	GPT-5.6 LunaOpenAISingle-shot multimodal · max reasoningPublished margin ±3.8 · 1,499 reported samples	58.10%	Not reported	Unknown
27	DeepSeek-V4.1-FlashDeepSeekSingle-shot multimodal · max reasoningPublished margin ±3.8 · 1,500 reported samples	56.70%	Not reported	Unknown
28	GPT-6 LunaOpenAISingle-shot multimodal · max reasoningPublished margin ±3.9 · 1,500 reported samples	55.50%	Not reported	Unknown
29	Sonnet 5AnthropicSingle-shot multimodal · high reasoningPublished margin ±3.7 · 1,500 reported samples	55.30%	Not reported	Unknown
30	Kimi K2.7 CodeKimiSingle-shot multimodal · high reasoningPublished margin ±3.7 · 1,500 reported samples	51.50%	Not reported	Unknown
31	Opus 4.7AnthropicSingle-shot multimodal · max reasoningPublished margin ±3.9 · 1,500 reported samples	46.90%	Not reported	Unknown
32	Sonnet 4.6AnthropicSingle-shot multimodal · high reasoningPublished margin ±3.9 · 1,500 reported samples	46.30%	Not reported	Unknown
33	MiniMax-M3MiniMaxSingle-shot multimodal · high reasoningPublished margin ±3.5 · 1,500 reported samples	41.70%	Not reported	Unknown

How to read this Mercor evaluation

We captured Mercor's SUPERChem results on October 15, 2026. The table preserves 33 published configurations, their source identifiers, reasoning settings, and disclosed pass@1 values. These are results from Mercor's evaluation setup; we did not rerun them.

This dated source table is display only and does not enter overall or category rankings. Source configurations retain their external identity, including Pro descriptors and different effort settings.

The headline is the source page's selected Pass@1 metric. The task set, harness, tools, and grader belong to this Mercor run and may differ from another evaluator's results.

The row labels retain published error margins and sample counts when supplied. Missing margins and counts remain unreported; a sample count is not used to reconstruct a task denominator. Except where the dedicated source defines an interval, the confidence level is unreported. Equal displayed scores and overlapping margins do not establish statistical ties. Capture time does not establish individual evaluation dates.

The general methodology lists programmatic multiple-choice answer extraction and a judge of None. Its run-count label is k = 3. The linked result page takes precedence where its setup differs, as noted above.

Snapshot

Pass@133 configurationsMercor evaluationDisplay only

Mercor SUPERChem results Mercor APEX benchmark index Mercor methodology Source paper Upstream evaluation code Upstream dataset

The published SUPERChem (Mercor) snapshot places GPT-6 Astra first at 80.30%. The third row is 2.50 points behind. The broader top-10 range is 7.50 points, so many of the published results sit in a relatively narrow band.

33 configurations are shown for SUPERChem (Mercor). The benchmark falls in the Knowledge category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. SUPERChem (Mercor) is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About SUPERChem (Mercor)

Tasks

500 source-reported tasks

Format

Pass@1

Difficulty

Source-specific evaluation

The selected headline is Pass@1. Source model identifiers, effort settings, published margins, and reported sample counts remain attached to the source configuration. Public-set results and Mercor task extensions occupy separate tables.

SUPERChem evaluation results Public benchmark source

Freshness and provenance

Version

SUPERChem; Mercor Pass@1

Refresh cadence

Reviewed dated source capture

Staleness state

Dated Mercor configurations

Question availability

Upstream task access and terms; only aggregate results mirrored

Dated Mercor configurationsDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does SUPERChem (Mercor) measure?

Answering advanced chemistry questions. This table shows Mercor-run configurations for reference and is excluded from model rankings.

Which model leads the published SUPERChem (Mercor) snapshot?

GPT-6 Astra currently leads the published SUPERChem (Mercor) snapshot with 80.30% pass@1. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on SUPERChem (Mercor)?

The October 15, 2026 capture snapshot contains 33 source configurations.

Last updated: October 15, 2026 capture · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 5,600+ readers.

One email each week. Unsubscribe anytime.