Skip to main content
BenchLM
Advertise

Model comparison

ZAYA1-74B-Preview vs ZAYA1-8B

Data verified

Head-to-head evidence from 5 shared benchmark results across 3 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

Sibling matchup inside the ZAYA1 family.

N/A
No comparison
ZAYA1-8B
Zyphra
N/A
0 category wins2 category wins

Evidence parity. ZAYA1-74B-Preview and ZAYA1-8B share 5 comparable benchmark results. 2 of 8 categories are comparable. 2 results are unique to ZAYA1-74B-Preview; 6 to ZAYA1-8B.

Updated July 23, 2026
Shared results
5
ZAYA1-74B-Preview only
2
ZAYA1-8B only
6
Comparable categories
2 / 8

ZAYA1-74B-Preview makes more sense if you need the larger 256K context window, while ZAYA1-8B is the cleaner fit if knowledge is the priority.

Confidence note. This is a partial-evidence comparison with 5 shared benchmark results across 3 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

ZAYA1-74B-Preview and ZAYA1-8B sit in the same ZAYA1 family. This page is less about two unrelated model lineages and more about how the siblings trade off on benchmark shape, token costs, and practical limits like context window.

ZAYA1-74B-Preview and ZAYA1-8B finish on the same BenchAlign overall score, so this is less about a single winner and more about where the edge shows up. The BenchAlign headline says tie; the benchmark table is where the real choice happens.

ZAYA1-74B-Preview gives you the larger context window at 256K, compared with 131K for ZAYA1-8B.

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 ZAYA1-74B-Preview and ZAYA1-8B
CategoryZAYA1-74B-PreviewΔZAYA1-8B
KnowledgeZAYA1-74B-Preview66.1Margin→ 7.5ZAYA1-8B73.6
MathZAYA1-74B-Preview76.4Margin→ 4.0ZAYA1-8B80.4
CodingZAYA1-74B-Preview53.2MarginNo overlapZAYA1-8BNot measured
Inst. FollowingZAYA1-74B-PreviewNot measuredMarginNo overlapZAYA1-8B64.1

Decisive benchmark drivers

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

More
A · ZAYA1-74B-PreviewB · ZAYA1-8B

  1. GPQA

    Knowledge
    Source ↗
    A 57.3%B 71%
    Winner: ZAYA1-8BΔ 13.7
    GPQA: ZAYA1-74B-Preview scored 57.3%; ZAYA1-8B scored 71%. ZAYA1-8B wins this benchmark.

  2. AIME26

    Math
    Source ↗
    A 76.4%B 89.1%
    Winner: ZAYA1-8BΔ 12.7
    AIME26: ZAYA1-74B-Preview scored 76.4%; ZAYA1-8B scored 89.1%. ZAYA1-8B wins this benchmark.

  3. MMLU-Pro

    Knowledge
    Source ↗
    A 68.1%B 74.2%
    Winner: ZAYA1-8BΔ 6.1
    MMLU-Pro: ZAYA1-74B-Preview scored 68.1%; ZAYA1-8B scored 74.2%. ZAYA1-8B wins this benchmark.

Operational comparison

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

MetricZAYA1-74B-PreviewZAYA1-8BComparison
Input / output priceUSD per 1M tokensZAYA1-74B-Preview$0 input / $0 outputZAYA1-8B$0 input / $0 outputListed prices are equal.
Generation speedtokens per secondZAYA1-74B-PreviewNot availableZAYA1-8BNot availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenZAYA1-74B-PreviewNot availableZAYA1-8BNot availableA complete latency comparison is not available.
Context windowmaximum listed tokensZAYA1-74B-Preview256KZAYA1-8B131KZAYA1-74B-Preview lists the larger context window.

Benchmark Deep Dive

Agentic
BenchmarkZAYA1-74B-PreviewZAYA1-8BResult
τ²-bench AirlineSource 56.1%—Not comparable
BFCL v4Source —39.2%Not comparable
Coding
BenchmarkZAYA1-74B-PreviewZAYA1-8BResult
LiveCodeBench v6Source 65.7%65.8%ZAYA1-8B leads
SWE-bench VerifiedSource 53.2%—Not comparable
KnowledgeZAYA1-8B wins
BenchmarkZAYA1-74B-PreviewZAYA1-8BResult
MMLU-ProSource 68.1%74.2%ZAYA1-8B leads
GPQASource 57.3%71%ZAYA1-8B leads
GPQA-DSource 57.3%71.0%ZAYA1-8B leads
MathZAYA1-8B wins
BenchmarkZAYA1-74B-PreviewZAYA1-8BResult
AIME26Source 76.4%89.1%ZAYA1-8B leads
HMMT Feb 2026Source —71.6%Not comparable
IMOAnswerBenchSource —59.3%Not comparable
ApexSource —32.2%Not comparable
Inst. Following
BenchmarkZAYA1-74B-PreviewZAYA1-8BResult
IFEvalSource —85.6%Not comparable
IFBenchSource —52.6%Not comparable
Frequently Asked Questions (3)

Which is better, ZAYA1-74B-Preview or ZAYA1-8B?

ZAYA1-74B-Preview and ZAYA1-8B are sibling variants in the ZAYA1 family, so the right pick depends on whether you value the better benchmark line, cheaper tokens, or the larger context window. They are tied on the BenchAlign leaderboard on the current data.

Which is better for knowledge tasks, ZAYA1-74B-Preview or ZAYA1-8B?

ZAYA1-8B has the edge for knowledge tasks in this comparison, averaging 73.6 versus 66.1. Inside this category, GPQA is the benchmark that creates the most daylight between them.

Which is better for math, ZAYA1-74B-Preview or ZAYA1-8B?

ZAYA1-8B has the edge for math in this comparison, averaging 80.4 versus 76.4. Inside this category, AIME26 is the benchmark that creates the most daylight between them.

Related Comparisons

Explore More

Zyphra Compare Pricing Methodology Find Your Best LLM Overall Rankings
Last updated: July 23, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.