En 30 secondes
Theo Browne says Ox Alpha, revealed as GLM-5.3-Flash, has become one of his most-used models. He praises its reported performance for its price, multimodal inputs and deployment on Chinese AI chips.
- 1Ox Alpha’s identityTheo says one of his most-used models turned out to be GLM-5.3-Flash.
- 2Performance and priceHe estimates performance comparable to Opus 4.8 at one tenth to one hundredth of the price; this is his assessment rather than a universal benchmark.
- 3Multimodality and chipsTheo highlights image, audio and video inputs and says the model runs on Chinese AI chips.
À retenir
- Ox Alpha was revealed as GLM-5.3-Flash.
- Theo considers its performance competitive with Opus 4.8 at a much lower price, based on his experience.
- The model accepts image, audio and video inputs and was presented as running on Chinese AI chips.
Lire le texte remanié, chapitre par chapitre
A fast, multimodal model
Theo shares his impressions of GLM-5.3-Flash, previously previewed as Ox Alpha. He discusses performance, pricing, inputs and the hardware used to run it.
Ox Alpha has quickly become one of my most-used models, and I really like using it. The surprise is that it is a Flash model: GLM-5.3-Flash.
In my experience, it behaves on par with models such as Opus 4.8 at one tenth to one hundredth of the price. That is my comparison, not a guarantee that the models perform equally on every task or at every provider’s prices.
The model can take images, audio and video. That stands out to me because I say the non-Flash GLM-5.3 could not even accept a screenshot. The final surprise: the announcement says the Ox Alpha preview ran entirely on Chinese AI chips.
Sur la même chaîne
38:47EN
Theo - t3․gg· 28 oct. 2026Theo’s Long-Running Agent Experiments: Verification, Compute and CostTheo Browne · Ping LabsA skateboard game and two compiler projects show how repeated checks and flexible compute can sustain agent work—and how expensive those loops can become.
36:07ENTheo - t3․gg· 27 oct. 2026Theo on Token Costs, Coding Agents, and the Price of AI CapabilityTheo Browne · Ping LabsA viral estimate of Lauren's AI spending prompts a closer look at cache pricing, benchmarks, and the changing work of coding agents.
30:37ENTheo - t3․gg· 26 oct. 2026Theo’s Three Skills for Reviewing and Shipping with Coding AgentsTheo Browne · Ping LabsHow sol-loop, full-send and full-stop encode review, verification and different levels of merge authority.
0:43ENTheo - t3․gg· 23 oct. 2026Theo Distinguishes Security Vulnerabilities from Active AttacksTheo Browne · Ping LabsA short explanation of why disclosed software flaws and reported attacks are different kinds of security news.
29:22ENTheo - t3․gg· 22 oct. 2026Theo on Security Vulnerabilities, Stolen API Keys and AITheo Browne · Ping LabsA look at the gap between vulnerability disclosure and patching, and a speculative account of why stolen inference keys may be valuable.
31:13ENTheo - t3․gg· 20 oct. 2026Theo on Why Smarter Models Help More Than DevelopersTheo Browne · Ping LabsFrom game modifications to compiler experiments, Theo revisits his assumptions about who benefits from more capable AI.