Skip to main content

Model comparison

MAI-Thinking-1 vs ZAYA1-8B

Data verified

Head-to-head evidence from 6 shared benchmark results across 3 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

N/A
No comparison
Zyphra
N/A
2 category wins1 category wins

Evidence parity. MAI-Thinking-1 and ZAYA1-8B share 6 comparable benchmark results. 3 of 8 categories are comparable. 7 results are unique to MAI-Thinking-1; 5 to ZAYA1-8B.

Updated July 23, 2026
Shared results
6
MAI-Thinking-1 only
7
ZAYA1-8B only
5
Comparable categories
3 / 8

Treat this as a split decision. MAI-Thinking-1 makes more sense if instruction following is the priority or you need the larger 256K context window; ZAYA1-8B is the better fit if knowledge is the priority.

Confidence note. This is a partial-evidence comparison with 6 shared benchmark results across 3 evidence categories; 3 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

MAI-Thinking-1 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.

MAI-Thinking-1 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 MAI-Thinking-1 and ZAYA1-8B
CategoryMAI-Thinking-1ΔZAYA1-8B
Inst. FollowingMAI-Thinking-185.0Margin 20.9ZAYA1-8B64.1
MathMAI-Thinking-189.7Margin 9.3ZAYA1-8B80.4
KnowledgeMAI-Thinking-172.5Margin 1.1ZAYA1-8B73.6
AgenticMAI-Thinking-146.0MarginNo overlapZAYA1-8BNot measured
CodingMAI-Thinking-165.5MarginNo overlapZAYA1-8BNot measured

Decisive benchmark drivers

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

More
A · MAI-Thinking-1B · ZAYA1-8B
  1. IFBench

    Inst. Following
    Source ↗
    A 85%B 52.6%
    Winner: MAI-Thinking-1Δ 32.4
    IFBench: MAI-Thinking-1 scored 85%; ZAYA1-8B scored 52.6%. MAI-Thinking-1 wins this benchmark.
  2. HMMT Feb 2026

    Math
    Source ↗
    A 84.9%B 71.6%
    Winner: MAI-Thinking-1Δ 13.3
    HMMT Feb 2026: MAI-Thinking-1 scored 84.9%; ZAYA1-8B scored 71.6%. MAI-Thinking-1 wins this benchmark.
  3. GPQA

    Knowledge
    Source ↗
    A 84.2%B 71%
    Winner: MAI-Thinking-1Δ 13.2
    GPQA: MAI-Thinking-1 scored 84.2%; ZAYA1-8B scored 71%. MAI-Thinking-1 wins this benchmark.
  4. MMLU-Pro

    Knowledge
    Source ↗
    A 85%B 74.2%
    Winner: MAI-Thinking-1Δ 10.8
    MMLU-Pro: MAI-Thinking-1 scored 85%; ZAYA1-8B scored 74.2%. MAI-Thinking-1 wins this benchmark.
  5. AIME26

    Math
    Source ↗
    A 94.5%B 89.1%
    Winner: MAI-Thinking-1Δ 5.4
    AIME26: MAI-Thinking-1 scored 94.5%; ZAYA1-8B scored 89.1%. MAI-Thinking-1 wins this benchmark.

Operational comparison

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

MetricMAI-Thinking-1ZAYA1-8BComparison
Input / output priceUSD per 1M tokensMAI-Thinking-1Not availableZAYA1-8B$0 input / $0 outputA complete price comparison is not available.
Generation speedtokens per secondMAI-Thinking-1Not availableZAYA1-8BNot availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenMAI-Thinking-1Not availableZAYA1-8BNot availableA complete latency comparison is not available.
Context windowmaximum listed tokensMAI-Thinking-1256KZAYA1-8B131KMAI-Thinking-1 lists the larger context window.

Benchmark Deep Dive

Agentic
BenchmarkMAI-Thinking-1ZAYA1-8BResult
Terminal-Bench 2.0Source 46%Not comparable
BFCL v4Source 39.2%Not comparable
Coding
BenchmarkMAI-Thinking-1ZAYA1-8BResult
SWE-bench VerifiedSource 73.5%Not comparable
SWE-bench ProSource 52.8%Not comparable
Terminal-Bench 2.0Source 46.0%Not comparable
LiveCodeBench v6Source 65.8%Not comparable
Reasoning
BenchmarkMAI-Thinking-1ZAYA1-8BResult
Graphwalks BFS 128KSource 90%Not comparable
KnowledgeZAYA1-8B wins
BenchmarkMAI-Thinking-1ZAYA1-8BResult
GPQASource 84.2%71%MAI-Thinking-1 leads
GPQA-DSource 84.2%71.0%MAI-Thinking-1 leads
MMLU-ProSource 85%74.2%MAI-Thinking-1 leads
SimpleQASource 31%Not comparable
MathMAI-Thinking-1 wins
BenchmarkMAI-Thinking-1ZAYA1-8BResult
AIME 2025Source 97%Not comparable
AIME26Source 94.5%89.1%MAI-Thinking-1 leads
HMMT Feb 2026Source 84.9%71.6%MAI-Thinking-1 leads
IMOAnswerBenchSource 59.3%Not comparable
ApexSource 32.2%Not comparable
Inst. FollowingMAI-Thinking-1 wins
BenchmarkMAI-Thinking-1ZAYA1-8BResult
IFBenchSource 85%52.6%MAI-Thinking-1 leads
IFEvalSource 85.6%Not comparable
Frequently Asked Questions (4)

Which is better, MAI-Thinking-1 or ZAYA1-8B?

MAI-Thinking-1 and ZAYA1-8B are tied on the BenchAlign overall score, so the right pick depends on which category matters most for your use case.

Which is better for knowledge tasks, MAI-Thinking-1 or ZAYA1-8B?

ZAYA1-8B has the edge for knowledge tasks in this comparison, averaging 73.6 versus 72.5. Inside this category, GPQA is the benchmark that creates the most daylight between them.

Which is better for math, MAI-Thinking-1 or ZAYA1-8B?

MAI-Thinking-1 has the edge for math in this comparison, averaging 89.7 versus 80.4. Inside this category, HMMT Feb 2026 is the benchmark that creates the most daylight between them.

Which is better for instruction following, MAI-Thinking-1 or ZAYA1-8B?

MAI-Thinking-1 has the edge for instruction following in this comparison, averaging 85 versus 64.1. Inside this category, IFBench is the benchmark that creates the most daylight between them.

Related Comparisons

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.