Model

Hy3

OPEN

Tencent · Hunyuan family · released Jul 6, 2026

Full release of Tencent's Hunyuan 3 (upgraded from an April 2026 preview build), Tencent's first genuinely unrestricted Apache 2.0 flagship. Newer than Hunyuan-A13B already in corpus (dated 2025-06-27). Pricing converted from RMB 1/RMB 4 per million input/output tokens at ~7.1 RMB/USD.

ReasoningCodingVisionFunction callingTool useAgentic

2460.0

Elo · rank #36

Parameters

295B

Active params

21B (MoE)

Context

256K tokens

Architecture

Mixture-of-Experts hybrid fast/slow-thinking model, 295B total / 21B active

License

Apache 2.0

Languages

—

API price (in/out)

$0.14 / $0.55

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

BrowseCompAgents84.2%#11

best: Kimi K3 · 91.2%

GPQA DiamondReasoning90.4%#28

best: GPT-6 Astra · 96.0%

Humanity's Last ExamReasoning37.0%#34

best: Claude Opus 5 · 64.7%

SWE-bench ProCoding57.9%#24

best: Claude Fable 5.1 · 81.2%

SWE-bench VerifiedCoding78.0%#28

best: Claude Opus 5 · 96.0%

Terminal-Bench 2.0Coding71.7%#24

best: Gemini 3.8 Flash · 89.4%

USAMOMath72.0%#3

best: Claude Opus 4.8 · 96.7%

Run it locally

VRAM @ Q4

—

VRAM @ FP16

—

Fits on (Q4)

Multi-node cluster required

Throughput data unavailable.

Quantizations

—

Fine-tune it

Permissive

QLoRA200.6 GB2× H200 141GB

LoRA628.4 GB4× B200 192GB

Full fine-tune4734.8 GBbeyond 8× B200

QLoRA SFT on ~10k samples ≈ $10.86 (2× H200 141GB)

Plan a fine-tune →

Hunyuan family

Elo progression across releases

Hunyuan (1st-gen)Sep 20231000.0 Hunyuan-Large (A52B)Nov 20241020.5 Hunyuan-TurboSFeb 20251267.4 Hunyuan-T1Mar 20251610.8 Hunyuan-A13BJun 20251513.6 Hy3Jul 20262460.0

Resources

HFtencent/Hy3 DOCShttps://www.tencent.com/en-us/articles/2202386.html APIAvailable via Tencent API

API price $0.14/$0.55 · each benchmark row carries its own source badge (see methodology)