GLM-5.3 Lanseras: Läckan Var Verklig — Z.ais post-tränade flaggskepp för kodning och cyberförsvar

The leak was real, the launch is here, and GLM-5.3 now has an independent score to argue about. Artificial Analysis' Intelligence Index — measured by the lab, not by Z.ai — puts GLM-5.3 at 60, tied with Kimi K3 for the top open-weights score on the board and 7 points clear of GLM-5.2's 53. The API went live this week at the same price as its predecessor, the open weights are confirmed for Friday, August 28, and the coding and cyber-defense claims Z.ai has been making since the August 14 announcement are starting to become testable. This page first tracked GLM-5.3 from its August 3 leak traces; this is the launch report, updated in place with what the launch, the API, and the first independent benchmark actually confirmed.

Från läcka till lansering

The four traces that surfaced on August 3 — a "ZCode for GLM-5.3" harness page, an official docs page reachable for roughly an hour, a Bing index entry reading "GLM-5.3 Official Harness," and a commit adding a "glm-5.3" entry with JSON Schema support to Zhipu's official Java SDK — all pointed at a real, named release. Z.ai co-founder Tang Jie's "sooooooon" reply and the "epic-level plus" framing are now confirmed by an actual product rather than a rumor. The "roughly a week" timing signal this page tested — posted on X by @teortaxesTex after DeepSeek V4 Pro shipped on August 13 — held to the day: Z.ai formally announced GLM-5.3 on August 14 under the slogan "Built to Code. Ready for Cyber Defense." The follow-on came this week: on August 19 Z.ai said the GLM-5.3 API was live and open for calls, priced the same as GLM-5.2, with the model already wired into ZCode, AutoClaw, and the GLM Coding Plan.

Vad post-träningen faktiskt gav

Det arkitektoniska påståendet är den del som är värd att slå fast, eftersom läckans mest högljudda detalj — att parametrarna överstiger en biljon — är fel. GLM-5.3 är inte en större modell. Z.ai säger att den återanvänder exakt samma 743B Mixture-of-Experts-bas som GLM-5.2 (cirka 40 miljarder aktiva parametrar per token), behåller samma 1M-token-kontextfönster och cirka 128K maxutdata, och får alla sina vinster från uppskalad post-träning: fler uppgiftsmiljöer med lång horisont, fler miljötyper och längre träningskörningar, byggd på IndexShare långkontext, SAO asynkron-RL och slime-ramverket med öppen källkod som redan producerade GLM-5.2. Den där formuleringen "ingen omträning, allt post-träning" är Z.ai:s egen, och den har inte genomgått någon oberoende granskning. Storleken är den andra rubriken: med 743B totala parametrar, varav endast cirka 40B aktiva per token, är GLM-5.3 lätt nog att självhosta på ett blygsamt kluster och billig nog att serva i stor volym — den "mindre, billigare, öppen"-positionering som kommentarerna på lanseringsdagen kopplade till den, och en skarp kontrast till de stängda frontier-flaggskeppen som den mäts mot.

Kodning och agenter: siffrorna Z.ai hävdar

Inom kodning rapporterar Z.ai — allt leverantörsrapporterat och ej återproducerat — Terminal-Bench 3.0 stiger från 4,6 till 28,3, vilket Z.ai kallar det bästa open-weights-resultatet på den testmiljön; DeepSWE v1.1 stiger från 46,2 till 66,9; SWE-Marathon ungefär fördubblas från 19,4 till 42,5; och Agents' Last Exam (CLI) stiger från 23,8 till 28,5. På Z.ai:s interna kod-benchmark uppnådde GLM-5.3 31,4 % vid ungefär 50 000 utdatatokens per uppgift med hög ansträngning, jämfört med Claude Opus 4.8:s 29,5 % vid cirka 120 000 tokens — Z.ai:s poäng är att GLM-5.3 når ett jämförbart resultat samtidigt som den förbrukar betydligt färre utdatatokens. Claude Fable 5 leder fortfarande den interna utvärderingen med 39,5 % vid maximal ansträngning, och Z.ai medger att GLM-5.3 fortfarande ligger efter GPT-5.6 Sol och Claude Fable 5 på flera svårare kodutvärderingar. Behandla alla dessa som leverantörssiffror tills en oberoende testmiljö återger dem.

Cyberförsvar: förmågan ingen såg komma

De faktiska cybersäkerhetssiffrorna är nyheten, och de är helt och hållet leverantörsrapporterade. På CyberGym, ett white-box-riktmärke för sårbarhetsupptäckt och validering, rapporterar Z.ai att GLM-5.3 når 84,5 %, upp från GLM-5.2:s 77,2 % och före Anthropics Mythos 5 (83,8 %) och GPT-5.6 Sol (83,6 %). På ExploitBench, som kräver både grundorsaksanalys och en fungerande exploit, mer än fördubblades GLM-5.3 från 24,4 % till 54,4 %, även om Mythos 5 (78,0 %) fortfarande ligger före. På ExploitGym rapporterar Z.ai 105 slutförda uppgifter inom en 2-timmarsbudget och 130 inom 6 timmar, jämfört med 29 och 39 för GLM-5.2 – återigen bakom Mythos 5 (181 och 247). Z.ai beskriver cyberförmågan som en emergent egenskap av skalad efterträning – "förmågan fortsatte att växa i takt med att träningen skalades", enligt företaget – snarare än ett medvetet mål.

Z.ai lägger till ett verklighetsbaserat påstående som komplement till riktmärkena: i tester med säkerhetsteam identifierade GLM-5.3 2 436 sårbarheter i 269 open source-projekt, varav 1 097 klassades som kritiska eller allvarliga, med det äldsta fyndet från 1981 och en genomsnittlig "livslängd" på 26,6 år. Det är den mest slående siffran i tillkännagivandet och även den minst oberoende verifierbara. Företaget har kombinerat kapaciteten med en Security Disclosure Ledger för samordnat offentliggörande, ett "trusted access"-program som begränsar känsliga cyberfunktioner till verifierade användare, och ett "Open Source Shield"-initiativ för att kontinuerligt granska viktiga open source-projekt.

Baslinjen GLM-5.3 var tvungen att slå.

GLM-5.2 is the reference point the whole story hangs on. It shipped in June 2026 as a 743B Mixture-of-Experts model with roughly 40B active parameters per token, a 1M-token context window, a 128K max output, an MIT license, and open weights on Hugging Face. Independently, Artificial Analysis' Intelligence Index puts GLM-5.2 at 53 — the highest open-weights score on the index until this week. GLM-5.3 now clears it by 7 points: Artificial Analysis measures GLM-5.3 at 60 on the same index (v4.1.1), tying Kimi K3 for the top open-weights position and landing it in the frontier band alongside closed flagships like Claude Fable 5 and GPT-5.6 Sol. That is the first independent number attached to GLM-5.3, and it is consistent with the direction — if not every detail — of Z.ai's own claims. On long-horizon coding, the OrcaRouter harness measures 77.9 on Terminal-Bench 2.1, while Z.ai's best-reported GLM-5.2 figure is 82.7, which would be the first open-weight score above 80 but is vendor-reported and unreproduced. The list price is $1.40 per million input and $4.40 per million output tokens.

The scoreboard above is the leak-era projection this page published before launch — the ">1T params (rumored)" row, the unconfirmed context and license, the projected AA index. The launch corrected the biggest cell: GLM-5.3 reuses the same 743B base as GLM-5.2, so there is no parameter jump. The context window is confirmed at 1M, and the license stays unconfirmed because the open weights have not shipped yet. The projected index cell — this page's own guess of ~57–60 — was the rare projection that came in on the nose: the real number is 60, and the open question now is what happens when that score is reproduced against the actual weights.

The capture above is the independent baseline GLM-5.3's claims are measured against. GLM-5.2 tops the open-weights leaderboard at an Artificial Analysis Intelligence Index of 53. The first test of whether GLM-5.3's post-training deltas move that number has now arrived: Artificial Analysis measures GLM-5.3 at 60 on the same index — tied with Kimi K3 for the open-weights lead, 7 points ahead of GLM-5.2, and reported by the lab as independently measured.

Priser och tillgänglighet

GLM-5.3 is priced identically to GLM-5.2: $1.40 per million input and $4.40 per million output tokens (¥8 / ¥28 in the domestic listing), with cached-input reads at $0.26 / ¥2 per million. Z.ai announced the API was open on August 19, and it is reachable through Z.ai's own API, ZCode, AutoClaw, the GLM Coding Plan, and several partner gateways. One behavior change matters for API callers: requests now require "thinking" enabled across three effort levels — low, high, and max — with no off switch, a breaking change for existing integrations.

Same price does not mean same bill. GLM-5.3 runs roughly 20% more tokens per task than GLM-5.2 did on the same workloads, which a cost-per-task reading puts at about $0.68 against GLM-5.2's $0.44 — still under Kimi K3 (about $0.84) and GPT-5.6 Sol (about $1.23). That per-task math is a derived estimate from observed token usage, not a vendor figure, but it is the number that decides whether the flat $1.40 / $4.40 rate card actually saves you money.

Vad lanseringen förändrar för dig

For API callers already on GLM-5.2, the practical step is a model-name change, not a project: GLM-5.2 is OpenAI-compatible and the integration carries over, with the thinking-effort caveat above. For self-hosters, the timeline is now a date rather than a guess: Zhipu promised the weights "two weeks after release" on August 14, which lands on Friday, August 28, and the open question is whether the license stays permissive. For anyone comparing models in the DeepSeek V4 Pro, Qwen3.8-Max, Kimi K3, GPT-5.6 Sol, and Claude Fable 5 tier, GLM-5.3 is now a live, independently scored variable in that ranking instead of a rumor.

The launch-day argument around GLM-5.3 is that the coding frontier has converged: for most everyday tasks, the story goes, few users can reliably tell GPT-5.6 Sol, Claude Fable 5, Kimi K3, GLM-5.2, and Qwen3.8-Max apart. If that convergence is real, the deciding factors stop being raw capability and become price, openness, and switching cost — which is exactly the corner GLM-5.3 is staking out at $1.40 / $4.40 per million on a self-hostable 743B base with weights confirmed for August 28. Whether coding models are genuinely interchangeable is an opinion, not a benchmark; the prices, the parameter count, and the weight date are not.

On the routing side, GLM-5.3 went live on OrcaRouter on August 18, the same day Z.ai's API opened — at the first-party list price, $1.40 / $4.40 per million, passed through with zero markup. The screenshot below shows GLM-5.2's page, which is exactly the shape GLM-5.3 now has: same price, same 1M-token context, same 128K max output. Routing a slice of real traffic to GLM-5.3 with automatic failover to GLM-5.2 or another proven model is a configuration change, not a rewrite — same key, no second contract. If the new model regresses on your workload, the router falls back before a page turns, and you get a quality signal on your own traffic instead of a vendor's slide. For a model whose flagship claims are still mostly vendor-reported, that is the low-risk way to find out for yourself.

Vad du ska titta på härnäst

• The weights, on Friday, August 28, and the license line on the model card — permissive MIT like GLM-5.2, or something narrower. Zhipu's cyber-safety hardening is the stated reason for the two-week delay, and the "trusted access" program suggests some functions will be gated regardless.

• Whether the cyber claims hold up outside Z.ai's own harness. The 2,436-vulnerability real-world claim and the CyberGym lead are the numbers independent labs will probe first; the AA Intelligence Index measures general capability, not security.

• Where the index lands once the weights are out. The 60 is scored against the served API; the self-hosted version, with a license attached, is the one teams will actually redeploy.

• DeepSeek V4 Flash's announced price increase, which sets the pricing envelope GLM-5.3 is being judged against, and GPT-5.6 Sol's one-point lead at 61.

FAQ

Är GLM-5.3 en större modell än GLM-5.2?

Nej. Z.ai säger att GLM-5.3 använder samma 743-miljarder-parameters Mixture-of-Experts-bas som GLM-5.2, med samma 1M-token-kontextfönster och ungefär 40 miljarder aktiva parametrar per token. Alla rapporterade förbättringar kommer från skalad efterträning, inte en parameterökning — vilket direkt korrigerar ryktet från läckageeran om en bas över en biljon. Det arkitekturpåståendet är Z.ai:s eget och har inte oberoende granskats.

När kommer GLM-5.3:s öppna vikter att vara tillgängliga?

Friday, August 28. Zhipu promised the weights "two weeks after release" when it announced GLM-5.3 on August 14, and said "next Friday" when the API went live on August 19 — both readings land on the same date. The license has not been confirmed, and Z.ai has said sensitive cyber functions will be restricted to a verified-user "trusted access" program.

Hur ska jag behandla benchmark-siffrorna?

Split the list. The coding jumps — Terminal-Bench 3.0 at 28.3, DeepSWE v1.1 at 66.9, SWE-Marathon at 42.5 — and the cyber results — CyberGym 84.5%, ExploitBench 54.4% — all come from Z.ai's own announcement and remain vendor-reported until an independent harness reproduces them. The Artificial Analysis Intelligence Index of 60 is the first independent measurement, and it is the number to weigh against everything Z.ai claims.

The leak was real, and the launch confirmed the name, the framing, and the timing — while correcting the one specific the rumor mill got loudest about. GLM-5.3 is the same base, post-trained hard, and now it carries an independent score to hold its vendor claims against: 60 on the Artificial Analysis Intelligence Index, tied with Kimi K3, seven ahead of GLM-5.2. The coding and cyber numbers are still Z.ai's own, the weights land on August 28, and the license line and a genuinely independent probe of the security claims are what's left to settle. Until then, the low-risk way to form your own view is a slice of real traffic and a failover to something proven.