Topic
Per-token rates, subscription quotas and cost-per-task comparisons across frontier and open-weight providers.
TypeSafe now fronts every new account $5 of credit it values at roughly 120 million tokens. Decision-model calls are far smaller than LLM completions, so that allowance does not convert into a request count.
Reality
GLM-5.3-Flash leads agentic terminal work, DeepSeek V4 Flash is billed as the cheapest per token, and a 2.52B MiniCPM5-2B runs locally under Apache 2.0. The comparison flags most of those numbers as vendor-reported.
Reality
A Guardian column by Heather Stewart argues the nearer risk in AI is financial, counting $132bn of hyperscaler debt issuance this year and a $1.5tn compute bill that lands within about two years.
Reality
A team that now has access to the model behind a published 193.6x figure measured 12x at the median on its own admin workload, and found that 83 percent of its bill was output tokens the new model does not charge for.
Reality
Z.ai's 320-billion-parameter model activates 18 billion per token and ships under MIT, so a buyer can download it and measure for themselves. Every capability figure published so far comes from Z.ai's own launch materials.
Reality
AWS has put Moonshot AI's 2.8-trillion-parameter model behind Bedrock's APIs and data boundary. The explicit prompt caching it ships with only pays back if you reuse a prefix inside half an hour.
Reality
The langchain-typesafe package lets an agent submit its state and a list of pre-defined questions in one request, and the speed figures behind it are TypeSafe's own, measured from laptops beside its own service.
Reality
TypeSafe's first model, Jev, answers structured questions with typed output and a confidence measure attached. The account of its launch says developers have to check those scores against real outcomes on their own data first.
Reality
The calculator prices its featured coding workload at about $6.94 a month through OpenRouter against 26 cents of electricity. The machine is still $3,419 down after a year and 7.97 billion tokens from break-even.
Reality
The rate undercuts the cached-input prices OpenAI and Anthropic publish by more than 130 times, and DeepSeek's own release says the sparse-attention design behind it has untested limits at cache boundaries.
Reality
Cost per request is down 34% at Uber and cost per session down 52%, and total AI spend has been flat since March even as token use grew. Pinterest and AT&T report similar savings from open models.
Reality
The MIT-licensed 320B model card claims it beats GLM-5.2 at a tenth of the price and approaches Claude Opus 4.8 on coding, but it names no dollar rate, and the comparisons are largely the vendor's own.
Reality
Speaker turns and end-of-speech markers now ride in the same token sequence as the words, which buys a cheaper pipeline and a tighter coupling than a three-vendor speech stack. The accuracy claim behind it is English-only.
Reality
Zhipu's open-weight MoE ties Claude Opus 4.8 on one composite index at roughly a fortieth of its list rate. At that price gap, the practical question stops being which model is better. It becomes how many attempts your router can afford.
Reality
One team's forty lines of Go moved 81% of requests to a cheaper model and cut the bill 71%, but the arithmetic says the win came from a boring traffic mix and a wide price gap, and the bill it pays is tail latency.
Reality
The share buying model-serving platforms went from 4.5% in January to 6.1% in July, and two named buyers have now moved production work onto Chinese open weights while incumbent revenue has not yet moved.
Reality
Artificial Analysis puts Grok 4.6 at 61 on its Intelligence Index, level with GPT-5.6 Sol, at $2/$6 per million tokens. The same pages record 48 seconds to first token.
Reality
Seven days after launch, xAI's flagship sits inside AWS procurement with a 500K context and four reasoning tiers. The rate card is flat; the effort dial is where the cost moves.
Reality
MAI-Thinking-1 is in public preview in Microsoft Foundry at $2 per million input tokens. For C# shops the consequence is an interface swap, not a new runtime to operate.
Reality
Z.ai says its 743B-parameter GLM-5.3 hits 34.5% on its own code bench using 22% fewer output tokens than GLM-5.2. The weights are still two weeks out.
Reality