Models / Hy4 Preview
Hy4 Preview
Early evidence, 4 results from 1 source, on 0 use cases. Hy4 Preview has not been measured on any use case yet; there is too little evidence yet to recommend it either way.
Results 16 Nov 2026 · 4 results · 1 source
Test Hy4 Preview on your prompt → It runs beside the models you could use instead, with every answer checked and costed.
- Price per million tokens
- $0.83 in · $2.50 out
Typical host on OpenRouter, 10 Nov 2026 · compare prices - Speed
- Not measured
- Intelligence Index
- Not measured
- Spring Prompt overall
- Not ranked yet
needs our benchmarks and two groups of results - Developer
- Tencent, based in China
Weights published - Released
- 28 Aug 2026
1,048,576 tokens of context
Where it stands among the models you could choose
Available models grouped by price, intelligence, speed and developer location, plus models with published weights. Retired models are excluded.
| Among | Intelligence Index | Spring Prompt overall | Price, blended | Speed |
|---|---|---|---|---|
| Models available today | – | – | $1.25 blended, 61st percentilecheapest: Mistral Nemo | – |
| Models from developers based in China | – | – | $1.25 blended, 79th percentilecheapest: Qwen3.5-9B | – |
| Open-weights models | – | – | $1.25 blended, 84th percentilecheapest: Mistral Nemo | – |
The name under each rank is the leader of that group; green is the top quarter of the group and red the bottom quarter, by rank. Price is shown as a percentile: the share of the group that costs less, so lower is cheaper. "Sets this group" marks the measure the group is defined by. Spring Prompt overall is our score out of 100 across every source (how it works). Price is the developer's list price (or the typical OpenRouter host where we haven't read one) for three input tokens to one output; intelligence and speed are from Artificial Analysis (data sourced from Artificial Analysis); "similar intelligence" means within 4 points on its Intelligence Index. Developer locations are where each company is based, not where a model is served. Groups under 3 models are not ranked.
How good is it, and for what?
Each use case ranks the available models its benchmarks measured, then those priced $1–3 per million tokens. The bar shows its percentile in that field, best to the right. Open a row for the results behind it.
Product listingsTurning a sparse product feed and photos into listings that can go live
- CatalogBench: has not measured this model.
Decks from an analysisTurning a finished analysis into a deck you could present as it is
- DeckBench: has not measured this model.
User surveysPlanning a user survey and reading its results without being misled
- SurveyBench: has not measured this model.
Marketing planningPlanning a year of ad spend without overspending
- ROASBench: has not measured this model.
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
- tau2-bench: pending. Its latest results are dated 26 May 2026, before this model was released.
- Berkeley Function Calling Leaderboard (BFCL) V4: pending. Its latest results are dated 16 Dec 2025, before this model was released.
- OpenHands Index: pending. Its latest results are dated 30 Jun 2026, before this model was released.
- Microsoft STATE-Bench: pending. Its latest results are dated 29 May 2026, before this model was released.
- Artificial Analysis: has not measured this model.
- Vending-Bench 2: has not measured this model.
Professional workReal tasks from banking, consulting and law, business documents and freelance projects
- APEX-Agents: has not measured this model.
- GDP.pdf: has not measured this model.
- Remote Labor Index: has not measured this model.
Reasoning and knowledgeHard questions across science, maths and general knowledge
- Artificial Analysis: has not measured this model.
WritingWhat people prefer in blind comparisons, and judged writing quality
- Arena (formerly LMArena): has not measured this model.
- UGI Leaderboard: has not measured this model.
Sticking to the factsSummarising without inventing things, and factual answers
- Vectara Hallucination Leaderboard: has not measured this model.
- Arena (formerly LMArena): has not measured this model.
- SimpleQA Verified (Epoch AI): has not measured this model.
Speed under pressureGood decisions against a real clock (fast chess)
- BulletBench: has not measured this model.
Where to run it
6 hosts on OpenRouter at the standard service tier, with a median of $1.25 blended per million tokens; quantisation where the host states it: FP8, FP4, BF16. Cheapest first.
| Host | Quantisation | Input | Output | Blended |
|---|---|---|---|---|
| AtlasCloud | FP8 | $0.83 | $2.50 | $1.25 |
| DeepInfra | FP4 | $0.83 | $2.50 | $1.25 |
| GMICloud | BF16 | $0.83 | $2.50 | $1.25 |
| Novita | FP8 | $0.83 | $2.50 | $1.25 |
| SiliconFlow | FP8 | $0.83 | $2.50 | $1.25 |
| Tencent | FP8 | $0.83 | $2.50 | $1.25 |
Prices per million tokens from OpenRouter's public endpoint list, checked 10 Nov 2026.
Against the alternatives
The models you would most likely weigh it against: the leaders of the groups above.
Scroll sideways to see every alternative.
| Hy4 Preview | Claude Fable 5.2 leads overall | Claude Opus 5.5 near the top overall | GPT-6.1 Sol near the top overall | GPT-6.1 Terra near the top overall | |
|---|---|---|---|---|---|
| Price per million tokens | $1.25 | $16.00 | $8.00 | $4.00 | $2.00 |
| Tokens a second | – | – | 96 | 53 | – |
| Intelligence Index | – | – | 57.6 | 51.8 | – |
| Spring Prompt overall | – | 90 | 82 | 80 | 80 |
| Rank among available models, by use case | |||||
| Product listings | – | 5th | 13th | 2nd | 4th |
| Decks from an analysis | – | 4th | 7th | 5th | 8th |
| User surveys | – | 1st | 4th | 3rd | 8th |
| Marketing planning | – | 1st | 9th | – | 2nd |
| Agents and tool use | pending | – | 3rd | 6th | 7th |
| Professional work | – | – | 3rd | 9th | – |
| Reasoning and knowledge | – | – | 1st | 6th | 22nd |
| Writing | – | – | 5th | 16th | – |
| Sticking to the facts | – | – | 3rd | 18th | – |
| Speed under pressure | – | 8th | – | 3rd | 17th |
Shaded figures are better than Hy4 Preview's. A dash means no result.
Where it has been measured
Not yet measured on any use case. Results for agents and tool use are pending: those sources last updated before this model was released.
Not timed: Artificial Analysis, our source for speed, has not tested it
- Arena (formerly LMArena)4 results
- Berkeley Function Calling Leaderboard (BFCL) V4Pending: latest results 16 Dec 2025
- Microsoft STATE-BenchPending: latest results 1 Jul 2026
- OpenHands IndexPending: latest results 30 Jun 2026
- tau2-benchPending: latest results 3 Aug 2026
- BulletBenchNot measured
- CatalogBenchNot measured
- DeckBenchNot measured
- ROASBenchNot measured
- SurveyBenchNot measured
- APEX-AgentsNot measured
- Artificial AnalysisNot measured
- GDP.pdfNot measured
- Remote Labor IndexNot measured
- SimpleQA Verified (Epoch AI)Not measured
- UGI LeaderboardNot measured
- Vectara Hallucination LeaderboardNot measured
- Vending-Bench 2Not measured
Every result
4 published results from 1 source, each in the source's own units, with the configuration that produced it.
Arena (formerly LMArena) · reported by Arena (formerly LMArena) · 4 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Arena Agent: confirmed task successIPS effect estimate, higher is better | 0.06 0.04–0.07 |
14 of 50 | Hy4 Preview (arena agent) | 2 Oct 2026 |
| Arena Agent: praise over complaintIPS effect estimate, higher is better | 0.07 0.04–0.10 |
18 of 50 | Hy4 Preview (arena agent) | 2 Oct 2026 |
| Arena Agent: steerabilityIPS effect estimate, higher is better | 0.01 -0.00–0.02 |
29 of 50 | Hy4 Preview (arena agent) | 2 Oct 2026 |
| Arena Agent: tool groundingIPS effect estimate, higher is better | 0.00 -0.00–0.00 |
33 of 50 | Hy4 Preview (arena agent) | 2 Oct 2026 |
Not shown: Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
Will Hy4 Preview work for your product?
Benchmarks aren't your task. Paste the prompt your product uses and run it beside the models you could use instead, with every answer checked and costed.
New here? Your email starts a free trial. No credit card needed.