Hunyuan Hy3 vs MAI-1-preview

Tencent · China  |  Microsoft · US · Updated June 2026

Quick verdict

Pick Hunyuan Hy3 for frontier-level reported reasoning and science (gpqa diamond 90.4) at low active-parameter cost or runs a 295b model at the cost of a 21b — only 21b parameters active per token. Pick MAI-1-preview for microsoft's first fully in-house, end-to-end foundation model - a historic break from relying solely on openai or ranked in the top 15 on lm arena at launch. Choose Hunyuan Hy3 if you need self-hosting or data privacy; MAI-1-preview if you want a managed API.

Hunyuan Hy3 (Tencent, China) and MAI-1-preview (Microsoft, US) line up two different AI ecosystems against each other — a comparison that is as much about cost philosophy and openness as raw capability. Hunyuan Hy3 is a 295B Apache-2.0 open MoE that reaches frontier reasoning quality while running at roughly 21B active-parameter cost. MAI-1-preview is microsoft's first fully in-house foundation model - a strategic break from sole reliance on OpenAI, trained on ~15,000 H100 GPUs. They diverge most on context window and open vs. closed weights — each quantified below from the models' real specs.

Key differences at a glance

Side-by-side specs

SpecHunyuan Hy3MAI-1-preview
ProviderTencent (China) Microsoft (US)
ReleasedJuly 6, 2026 August 28, 2025
Context window256K (~384 pages) 128K (~192 pages)
Price (in/out)Open weight (self-host / free) Not published
Open weight?Yes — self-hostable No — API only
Modalitiestext, code text, code
SWE-Bench VerifiedNot published Not published
MRCR v2 @ 1MNot published Not published

Who wins what

Frontier-level reported reasoning and science (GPQA Diamond 90.4) at low active-parameter cost

Hunyuan Hy3

MAI-1-preview is comparatively weak here — distinct from Microsoft's later MAI-Thinking-1 reasoning model - an earlier, more general-purpose entry

Runs a 295B model at the cost of a 21B — only 21B parameters active per token

Hunyuan Hy3

Its 256K window holds about 2× more than MAI-1-preview's 128K in a single prompt.

Clean, unrestricted Apache-2.0 license with no geographic carve-out

Hunyuan Hy3

A 295B Apache-2.0 open MoE that reaches frontier reasoning quality while running at roughly 21B active-parameter cost — and it carries the larger 256K context.

Microsoft's first fully in-house, end-to-end foundation model - a historic break from relying solely on OpenAI

MAI-1-preview

MAI-1-preview lists microsoft's first fully in-house, end-to-end foundation model - a historic break from relying solely on OpenAI among its strengths; Hunyuan Hy3 does not.

Ranked in the top 15 on LM Arena at launch

MAI-1-preview

MAI-1-preview lists ranked in the top 15 on LM Arena at launch among its strengths; Hunyuan Hy3 does not.

Trained on roughly 15,000 NVIDIA H100 GPUs, a genuine internal infrastructure investment

MAI-1-preview

MAI-1-preview lists trained on roughly 15,000 NVIDIA H100 GPUs, a genuine internal infrastructure investment among its strengths; Hunyuan Hy3 does not.

Largest single-prompt input

Hunyuan Hy3

Its 256K window is about 2× larger than MAI-1-preview's 128K, fitting roughly 384 pages in one prompt.

Which should you pick?

Someone analysing very long documents or codebases

Hunyuan Hy3

Larger 256K window fits more in one prompt.

A team with data-privacy or self-hosting needs

Hunyuan Hy3

Open weights let you run it on your own hardware; MAI-1-preview is API-only.

Anyone whose priority is frontier-level reported reasoning and science (gpqa diamond 90.4) at low active-parameter cost

Hunyuan Hy3

It is specifically built for that.

Anyone whose priority is microsoft's first fully in-house, end-to-end foundation model - a historic break from relying solely on openai

MAI-1-preview

That is its strongest area.

An enterprise with regional data-residency rules

MAI-1-preview or Hunyuan Hy3

Origin (China vs US) affects where data is processed and which compliance regime applies — check the provider's terms for your region.

Hunyuan Hy3: where it fits

A 295B Apache-2.0 open MoE that reaches frontier reasoning quality while running at roughly 21B active-parameter cost. Released July 6, 2026 by Tencent, it is built for frontier-level reported reasoning and science (GPQA Diamond 90.4) at low active-parameter cost, runs a 295B model at the cost of a 21B — only 21B parameters active per token, clean, unrestricted Apache-2.0 license with no geographic carve-out, and broad day-one ecosystem support plus an FP8 checkpoint.

Its trade-offs are real: benchmarks are largely self-reported, and the ultra-low hosted pricing is a limited promotion, and the hosted API is China-jurisdiction, and self-hosting a 295B MoE still needs serious hardware. As an open-weight model, its running cost is your own hardware rather than a per-token fee.

MAI-1-preview: where it fits

Microsoft's first fully in-house foundation model - a strategic break from sole reliance on OpenAI, trained on ~15,000 H100 GPUs. Released August 28, 2025 by Microsoft, it is built for microsoft's first fully in-house, end-to-end foundation model - a historic break from relying solely on OpenAI, ranked in the top 15 on LM Arena at launch, trained on roughly 15,000 NVIDIA H100 GPUs, a genuine internal infrastructure investment, and rolled into Copilot alongside OpenAI models, giving Microsoft a real second option.

Its trade-offs: a 'preview' release, not yet positioned as Microsoft's primary Copilot model, no public per-token API pricing - not sold as a standalone product, and distinct from Microsoft's later MAI-Thinking-1 reasoning model - an earlier, more general-purpose entry.

The bottom line for this matchup

The defining split here is open vs. closed. Hunyuan Hy3 gives you weights you control — self-host it, fine-tune it, keep data in-house, pay only for hardware. MAI-1-preview gives you a managed, always-updated API with no infrastructure to run. Teams with GPUs, privacy requirements, or huge volume often favour the open model; teams that want zero ops and the latest capabilities favour the closed one. Capability is close enough that this operational question, not the benchmark, usually decides it.

Want both Hunyuan Hy3 and MAI-1-preview without two subscriptions? LumiChats gives you these plus 40+ models under one ₹69/day pass (about $1/day) — draft with one, cross-check with the other.

See pricing

Frequently asked questions

Is Hunyuan Hy3 or MAI-1-preview better for coding?

Public SWE-Bench figures are not available for either model, so the honest test is your own repository — run an identical real bug through both. By design, Hunyuan Hy3 leans toward frontier-level reported reasoning and science (gpqa diamond 90.4) at low active-parameter cost while MAI-1-preview leans toward microsoft's first fully in-house, end-to-end foundation model - a historic break from relying solely on openai, and that positioning usually predicts which feels better on your codebase.

Which is cheaper, Hunyuan Hy3 or MAI-1-preview?

Hunyuan Hy3 is open-weight, so self-hosting means no per-token fee (you pay for hardware instead), while MAI-1-preview is API-metered at Not published. For most teams without GPUs, the API model is cheaper to start; at very high volume, self-hosting can win.

Which has the bigger context window?

Hunyuan Hy3 — 256K vs 128K, about 2× larger. Useful only if the model actually reasons over the full window, which not all do.

Can I use both Hunyuan Hy3 and MAI-1-preview together?

Yes — a multi-model platform like LumiChats gives you Hunyuan Hy3, MAI-1-preview and 40+ others under one ₹69/day pass (about $1/day), so you can draft with one and cross-check with the other instead of buying two subscriptions.

Which is newer, Hunyuan Hy3 or MAI-1-preview?

Hunyuan Hy3 — released July 6, 2026, about 10 months after MAI-1-preview.

Related comparisons

Specifications and benchmarks reflect publicly reported figures as of June 2026 and may change as providers release updates. Always verify on your own workload.