Skip to main content
Radar

Five or fewer confirmed AI changes, with original sources, on mornings when something changed.A free source-linked morning brief.

Start the free Radar Brief
BenchLM
Tencent logo
Model A
Hy4 preview

Tencent

79.16/100

Estimated · Public rank #7

90% interval 69.3–89.0

Hy4 preview vs Ling 3.0 Flash

Updated August 28, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

InclusionAI logo
Model B
Ling 3.0 Flash

InclusionAI

53.89/100

Estimated · Public rank #109

90% interval 42.4–65.4

Decision reading

Hy4 preview has the higher public score, 79.16 versus 53.89, and the 90% score intervals do not overlap.

9 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

  • Long documents

    Prompts that approach the documented context limit

    Hy4 preview

    Hy4 preview has the larger documented context window.

    Confidence: documented

Show secondary and unsupported calls

  • Coding work

    Code generation, repair, and software-engineering tasks

    Not enough matched evidence

    The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.

    Confidence: limited

  • Agentic work

    Tool use, computer use, and multi-step task completion

    Not enough matched evidence

    No shared weighted benchmark basis supports a winner.

    Confidence: limited

  • Chat turn cost

    1K fresh input + 500 output tokens

    Not enough matched evidence

    A complete comparable API-rate estimate is not available for both models.

    Confidence: listed-rates

  • Cache-heavy agent loop cost

    200K cached + 20K fresh input + 10K output tokens

    Not enough matched evidence

    A complete comparable API-rate estimate is not available for both models.

    Confidence: listed-rates

  • Repository review cost

    50K fresh input + 3K output tokens

    Not enough matched evidence

    A complete comparable API-rate estimate is not available for both models.

    Confidence: listed-rates

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

Shared results
9
Hy4 preview only
19
Ling 3.0 Flash only
8
Like-for-like categories
1 / 8

1 category uses different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Category results, on a stated basis

Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.

Knowledge

Like-for-like
Hy4 preview
60.4
Ling 3.0 Flash
31.1
Weighted basis
2 vs 2 rows
Reading
Hy4 preview leads

Coding

Directional only
Hy4 preview
65.7
Ling 3.0 Flash
47.1
Weighted basis
1 vs 2 rows
Reading
Directional only

Agentic

Not comparable
Hy4 preview
Not measured
Ling 3.0 Flash
72.2
Weighted basis
0 vs 1 rows
Reading
Not comparable

Reasoning

Not comparable
Hy4 preview
Not measured
Ling 3.0 Flash
Not measured
Weighted basis
0 vs 0 rows
Reading
Not comparable

Math

Not comparable
Hy4 preview
Not measured
Ling 3.0 Flash
90.1
Weighted basis
0 vs 2 rows
Reading
Not comparable

Multilingual

Not comparable
Hy4 preview
Not measured
Ling 3.0 Flash
Not measured
Weighted basis
0 vs 0 rows
Reading
Not comparable

Multimodal

Not comparable
Hy4 preview
66.2
Ling 3.0 Flash
Not measured
Weighted basis
1 vs 0 rows
Reading
Not comparable

Instruction following

Not comparable
Hy4 preview
Not measured
Ling 3.0 Flash
74.5
Weighted basis
0 vs 1 rows
Reading
Not comparable

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

Hy4 preview
Self-hosted; infrastructure cost varies
Fits in one request
Ling 3.0 Flash
API rate not published
Fits in one request

Hy4 preview has no comparable published API token rate. Ling 3.0 Flash has no comparable published API token rate.

Repository review

50K fresh input + 3K output tokens

Hy4 preview
Self-hosted; infrastructure cost varies
Fits in one request
Ling 3.0 Flash
API rate not published
Fits in one request

Hy4 preview has no comparable published API token rate. Ling 3.0 Flash has no comparable published API token rate.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Hy4 preview
Self-hosted; infrastructure cost varies
Fits in one request
Cached-input rate unavailable
Ling 3.0 Flash
API rate not published
Fits in one request
Cached-input rate unavailable

Hy4 preview has no comparable published API token rate. Ling 3.0 Flash has no comparable published API token rate.

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

Documented inputs

Hy4 preview

Not sourced

Ling 3.0 Flash

Not sourced

Documented outputs

Hy4 preview

Not sourced

Ling 3.0 Flash

Not sourced

Provider availability

Hy4 preview

Not sourced

Ling 3.0 Flash

Not sourced

Reasoning profile

Hy4 preview

Reasoning

Ling 3.0 Flash

Reasoning

Weight access

Hy4 preview

Open Weight

Ling 3.0 Flash

Open Weight

License

Hy4 preview

Open Weight

Ling 3.0 Flash

Open Weight

Release date

Hy4 preview

2026-08-28

Ling 3.0 Flash

2026-07-23

If you already use one of these models
Deployment change
The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal
Hy4 preview has the higher public score, 79.16 versus 53.89, and the 90% score intervals do not overlap.
Workload cost
A complete comparable API-rate estimate is not available for both models.
Context tradeoff
Hy4 preview has the larger documented window (1M).

Run the same representative tasks against both endpoints before changing production traffic.

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence36 rows

Agentic

  • Terminal-Bench 2.1

    Hy4 preview85.4%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • CyberGym

    Hy4 preview78.4%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • WideResearch

    Hy4 preview83.9%
    Source
    Ling 3.0 Flash73.6%
    Source

    Hy4 preview leads this result

  • DRACO

    Hy4 preview77.2%
    Source
    Ling 3.0 Flash70.4%
    Source

    Hy4 preview leads this result

  • MCP Atlas

    Hy4 preview83.7%
    Source
    Ling 3.0 Flash65.5%
    Source

    Hy4 preview leads this result

  • Toolathlon-Verified

    Hy4 preview74.1%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • APEX-Agents

    Hy4 preview37.1%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • skillsBench

    Hy4 preview62.9%
    Source
    Ling 3.0 Flash44.8%
    Source

    Hy4 preview leads this result

  • JobBench

    Hy4 preview61.7%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • Agents' Last Exam

    Hy4 preview22.8%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • AutomationBench

    Hy4 preview32.1%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • BankerToolBench

    Hy4 preview78.6%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • HLE w/ tools

    Hy4 preview55.4%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • BFCL v4

    Hy4 preview—
    Ling 3.0 Flash73.0%
    Source

    Not directly comparable

  • BrowseComp

    Hy4 preview—
    Ling 3.0 Flash72.2%
    Source

    Not directly comparable

Coding

  • Terminal-Bench 2.1

    Hy4 preview85.4%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • SWE-bench Pro

    Hy4 preview65.7%
    Source
    Ling 3.0 Flash56.6%
    Source

    Hy4 preview leads this result

  • SWE Multilingual

    Hy4 preview82.9%
    Source
    Ling 3.0 Flash72.4%
    Source

    Hy4 preview leads this result

  • deepSwe

    Hy4 preview64.3%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • NL2Repo

    Hy4 preview58.9%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • ProgramBench

    Hy4 preview17.5%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • PostTrain Bench

    Hy4 preview35.6%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • sweMarathon

    Hy4 preview31.9%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • LiveCodeBench v5

    Hy4 preview—
    Ling 3.0 Flash82.8%
    Source

    Not directly comparable

  • SciCode

    Hy4 preview—
    Ling 3.0 Flash41.2%
    Source

    Not directly comparable

Reasoning

  • CritPt

    Hy4 preview16.9%
    Source
    Ling 3.0 Flash—

    Not directly comparable

Knowledge

  • GPQA

    Hy4 preview92.3%
    Source
    Ling 3.0 Flash85.0%
    Source

    Hy4 preview leads this result

  • GPQA-D

    Hy4 preview92.3%
    Source
    Ling 3.0 Flash85.0%
    Source

    Hy4 preview leads this result

  • HLE

    Hy4 preview55.4%
    Source
    Ling 3.0 Flash22.7%
    Source

    Hy4 preview leads this result

  • HLE w/o tools

    Hy4 preview43.4%
    Source
    Ling 3.0 Flash—

    Not directly comparable

Math

  • Apex

    Hy4 preview74.2%
    Source
    Ling 3.0 Flash—

    Not directly comparable

  • AIME26

    Hy4 preview—
    Ling 3.0 Flash93.2%
    Source

    Not directly comparable

  • HMMT Feb 2026

    Hy4 preview—
    Ling 3.0 Flash87.0%
    Source

    Not directly comparable

  • IMOAnswerBench

    Hy4 preview—
    Ling 3.0 Flash83.7%
    Source

    Not directly comparable

Multimodal

  • OfficeQA Pro

    Hy4 preview66.2%
    Source
    Ling 3.0 Flash—

    Not directly comparable

Instruction following

  • IFBench

    Hy4 preview—
    Ling 3.0 Flash74.5%
    Source

    Not directly comparable

Frequently asked questions

Which is better, Hy4 preview or Ling 3.0 Flash?

Hy4 preview has the higher public score, 79.16 versus 53.89, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.

Which is better for coding, Hy4 preview or Ling 3.0 Flash?

The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, Hy4 preview or Ling 3.0 Flash?

The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.

Which costs less, Hy4 preview or Ling 3.0 Flash?

Both models do not have comparable published API token rates, so this page does not name a universal price winner.

Which has the larger context window, Hy4 preview or Ling 3.0 Flash?

Hy4 preview has the larger documented context window: 1M, compared with 262K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated August 28, 2026

Watch Hy4 preview vs Ling 3.0 Flash

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.