Skip to main content
BenchLM
Advertise

Sent every week

The model changes worth your attention

Rank moves, price changes, launches, and one clear decision note. One email every week.

Get the next weekly brief

Read the sample below first. Subscribe when the format earns the space in your inbox.

Join 2,000+ readers.

What arrives

Useful change, not a news firehose

Each issue is built for people choosing, buying, or operating language models.

Rank moves with the evidence attached

See who moved, what changed, and whether the public benchmark coverage supports the move.

Material API price changes

Track pricing that could change a production shortlist, without filling your inbox with every minor update.

Release and availability context

Separate a launch headline from the access limits, evidence gaps, and reasons to wait.

One decision note

Leave with a model to consider, a lower-cost alternative, or a clear reason not to switch yet.

Real sample issue

See the whole format

This is the editorial content of a sent issue, adapted from email to the web.

Weekly LLM benchmark digest

The frontier is outrunning the scorecards

GPT-5.6 became generally available on July 9. Kimi K3 arrived seven days later. In between, another frontier lab released a 975-billion-parameter open-weight model. The release cycle is compressing, and the next model can now arrive before the last one has enough independent evidence to compare cleanly.

What changed in seven days

GPT-5.6 SolAvailable in the API at $5 input and $30 output per million tokens. The family adds lower-cost Terra and Luna tiers, plus a multi-agent ultra mode.Kimi K3Available in Kimi products and the API with a 1M-token window. API pricing is $3 input and $15 output per million cache-miss tokens.

The limit

Kimi K3 was still unranked when this issue was sent. Its launch table mixed model and harness effects, and the promised weights were due by July 27. Official results were useful launch evidence, not a clean substitute for independent runs in a comparable setup.

What to expect next

Families, not single flagships

Labs can refresh capability tiers on separate schedules, which makes a model name less durable than its exact version and effort setting.

More agent-dependent scores

The host model, harness, reasoning budget, and parallelism increasingly determine the result together.

Shorter buying windows

Keep a small production fixture and compare cost per completed task. A general leaderboard can narrow the field; it cannot reproduce your stack.

Analysis worth opening

Kimi K3’s Official Benchmark TableWhat the launch results established, what they did not, and why the model remained unranked at send time.Thinking Machines Chose Open Weights FirstWhy a new frontier lab made its 975B-parameter first release portable, and where the hardware bill set a practical boundary.AI Coding Agents Need ReceiptsThe test a coding-agent ranking should pass before it tells anyone what to buy.

Archive copy reflects the rankings, prices, and availability stated when this issue was sent. Current pages may show newer evidence.

Past issues

Read before you subscribe

Every archived issue is dated. Its claims stay tied to the evidence available on that send date.

The frontier is outrunning the scorecardsKimi K3 followed GPT-5.6 Sol by seven days. Comparable evidence is the scarce input.GPT-5.6 arrives, locked behind a previewGPT-5.6 lands in a government-gated preview, Anthropic holds the top two spots, and GLM-5.2 enters the top five.The frontier moved again, twiceOpus 4.7 reaches general availability, Kimi 2.7 enters preview, and the gap between major releases keeps shrinking.

Sponsored editions

Paid partner briefs are labeled before the first partner link. Provider-run claims stay attributed to the sponsor.

The 158 ms call boundarySponsored by Celeris. A bounded look at routing, extraction, scoring, and query rewriting with Celeris-1.
Subscribe to the weekly brief