Forward Deployed Engineer

OpenRouter's Ori configures four coding agents in one command. OpenRouter adds Span-01, a decision model that checks AI agents. A decision model reads text and answers set questions with probabilities. Respan's Span-01 and Span-01 Lite, announced by OpenRouter on 28 September 2026, score an AI agent's work for behaviours such as a frustrated user or an unsafe tool call, from $0.02 per million tokens. 00:0008:32 A decision model is an AI model that reads text and answers set questions with probabilities. On 28 September 2026 OpenRouter, the service that routes developers' requests to AI models from many providers, posted that two new ones are live: Span-01 and Span-01 Lite, from the San Francisco start-up Respan. "Send a span and the behaviors you care about, and get back the probability each one is present," OpenRouter wrote, with examples such as whether the user is frustrated or whether a tool call is safe to run. Span-01 costs $0.02 per million input tokens, with output free, and Span-01 Lite is free. On Respan's own Behavior Benchmark, Span-01 scores an overall F1 of 0.843, ahead of OpenAI's GPT-6 Luna at 0.815 and TypeSafe's decision model Jev at 0.715. What is a decision model? A decision model "reads application state and answers typed questions with probabilities, never text", OpenRouter's documentation says, and "The model judges, code computes." A chat model writes an answer in words that a person or another program then has to interpret. A decision model returns numbers that software can act on straight away: route this ticket to payments, block this action, flag this conversation for a person. Every decision model on OpenRouter answers the same three kinds of question, evaluated in parallel within one request: | Question type | What it asks | What comes back | | Choice | Which option from a set fits | The chosen option, a probability for each option and a confidence | | Noul | Whether a condition holds | The probability that it does | | Score | Where something sits on ordered levels | A probability-weighted position, a probability per level and a confidence | OpenRouter's worked example sends Jev one support ticket, "My checkout page shows a blank screen after I click Pay. I have tried two browsers.", with three questions. Jev answers that it is a bug with probability 0.96, picks the payments team with probability 0.84 and rates it "Blocking revenue right now" with probability 0.99. The call used 476 input tokens and cost $0.000019992. TypeSafe launched Jev on 15 September 2026 as the first of what it calls System One models. "Think of Jev as a frontier-intelligence function call: unstructured state in, typed probabilistic decisions out," TypeSafe's founder Diogo Almeida wrote. How does Span-01 work? Span-01 reads a span of an AI agent's work, such as one exchange between a user and an assistant, and returns for each behaviour you describe in a sentence the probability that it is present, absent or not observable. "Span-01 applies behavior definitions it has never seen and returns probabilities in one forward pass," Respan's documentation says. Its launch post of 24 September 2026 puts it more briefly: "One true forward pass. No token-by-token generation." Respan's own example sends a user asking "Please connect me to a person." and an assistant replying "I will connect you to its support team.", with one behaviour to check: whether the user wants a human. Span-01 Lite returns a probability of 0.73 that the behaviour is present, 0.25 that it is absent and 0.02 that the span does not show it. Respan says it trained the model's general classification reasoning with RLAIF before specialising it for detecting behaviours in agent traces. OpenRouter describes Span-01 as "suited for evaluation, guardrails, and monitoring of LLM and agent outputs at scale". Span-01 is the higher-accuracy tier of the family and Span-01 Lite the free, lighter one. Respan's own API for the models is in early access with a waitlist; on OpenRouter both are listed with public prices. How much does a check cost? A check on a 500-token span costs $0.00001 on Span-01, or $10 for a million checks, at OpenRouter's listed price of $0.02 per million input tokens on 28 September 2026. Output is free on both Span-01 and Jev, because a decision model returns a handful of numbers, not paragraphs. | Model | Maker | Input, per million tokens | Output, per million tokens | | Span-01 | Respan | $0.02 | Free | | Span-01 Lite | Respan | Free | Free | | Jev 1.13 | TypeSafe | $0.042 | Free | | GPT-6 Luna | OpenAI | $0.10 | $0.50 | OpenRouter's pages, which track response times as a rolling figure, showed Span-01 answering in about half a second on 28 September 2026, Span-01 Lite a little faster, and Jev in about a quarter of a second. OpenRouter lists both Respan models as released on 26 September. How good is Span-01? On Respan's own Behavior Benchmark, Span-01 scores an overall F1 of 0.843, ahead of GPT-5.6 Terra at 0.837 and GPT-6 Luna at 0.815, and well ahead of Jev at 0.715, according to the chart in Respan's launch post of 25 September 2026. F1 combines how many of the real cases a model catches with how many of its flags are correct, on a scale from 0 to 1. | Model | Overall F1 | | Span-01 | 0.843 | | GPT-5.6 Terra | 0.837 | | GPT-6 Luna | 0.815 | | DeepSeek V4 Flash | 0.814 | | Span-01 Lite | 0.761 | | Sonnet 5 | 0.717 | | Jev | 0.715 | | Qwen3 235B | 0.678 | Respan's post claims Span-01 is "2x cheaper, 18% better than Jev". Both hold on the published numbers: Jev's input price is 2.1 times Span-01's, and 0.843 is 17.9% higher than 0.715. The same post says "700x cheaper, 4% better than GPT-6 Luna". On the chart Span-01's score is 3.4% higher than Luna's. OpenRouter's listed prices put Span-01's input at a fifth of Luna's, a gap of five times, with Span-01's output free against Luna's $0.50 per million tokens; the 700 times is Respan's own figure. Who is Respan? Respan, formerly Keywords AI, started at the University of Illinois, where co-founders Andy Li and Raymond Huang met as engineering students, and joined Y Combinator's Winter 2024 batch, according to its website. Its company page on Y Combinator lists it in San Francisco, founded in 2023, with 23 staff, and describes the business as "Self-driving observability, evals, and gateway for AI agents": tools that log, test and route the calls an agent makes. The same page says the platform handles more than 1 billion logs and 2 trillion tokens a month. What this means for AI agents. Decision models make it cheap to check every step an AI agent takes, because each answer is a probability that costs a fraction of a cent and arrives in about half a second. An agent that answers customers or runs code produces long traces. Checking each step with a chat model such as GPT-6 Luna costs five times as much per input token, plus $0.50 per million output tokens, and returns prose that code still has to parse. OpenRouter now carries two families built for the job: TypeSafe launched Jev on 15 September 2026, and OpenRouter lists Respan's two models from 26 September. Questions people ask. * What is a decision model? - A decision model is an AI model that reads a block of text, such as an app's current state or an agent's conversation, and answers set questions with probabilities. OpenRouter's documentation describes three question types: a choice from a defined list, whether a condition holds, returned as the probability of yes, and a score along ordered levels. Software then acts on the numbers. * What is Span-01? - Span-01 is a decision model from Respan, a San Francisco start-up formerly called Keywords AI. It reads a span of an AI agent's work and returns, for each behaviour described in a sentence, the probability that the behaviour is present, absent or not observable. It went live on OpenRouter with a lighter, free version, Span-01 Lite, announced on 28 September 2026. * How much does Span-01 cost? - On OpenRouter, Span-01 costs $0.02 per million input tokens and nothing for output, and Span-01 Lite is free. A check on a 500-token span therefore costs $0.00001 on Span-01, or $10 for a million checks. TypeSafe's Jev 1.13 costs $0.042 per million input tokens. Sources. The frontier brief. Models shipping, chains breaking, qubits holding. The whole frontier in one short read, written by the desks that cover it. Join the waitlist and get in before launch. More AI.

Omarchy, an Arch-based Linux distribution created by Ruby on Rails founder David Heinemeier Hansson, has secured $18.5 million in backing, including multiyear commitments and AI tokens. Patrons include OpenRouter, Four Technologies, DigitalOcean, Meta Superintelligence Labs, Anthropic, OpenAI, and Fireworks. The project has released version 4.0.3 and announced Omarchy M for Apple Silicon Macs. The Omacom Foundation reports that the distribution's latest version was built almost exclusively by AI agents. Three developers have been hired, including kernel developer Krzysztof Wilczyński. However, Omarchy faces significant criticism in the open-source community. Critics, including prominent developers Matthew Garrett and Jürgen Geuter, have accused the project of promoting right-wing politics and being "fundamentally incompatible with the goals of free software." A "Stop Omarchy" campaign has emerged opposing the distribution.

Stripe & OpenRouter: acquisition powering AI token economics. August 24, 2026 Stripe is buying AI platform OpenRouter for a reported US$8bn, pairing global payment systems with smart token routing to control AI unit economics The financial services company Stripe has announced that it has agreed to acquire OpenRouter, a leading AI model gateway and routing platform. OpenRouter helps businesses route and optimise AI token usage across more than 400 models from more than 80 providers. Over the past month or so, OpenRouter's annualised revenue grew around 15%, to US$160m, in part due to OpenAI offering discounts on model usage, according to a person with knowledge of the business who spoke with The Information. AI Magazine believe intelligence will be multi-model: no single model will be optimal for every task and developers need a neutral layer to orchestrate and manage them all OpenRouter's CEO Alex Atallah The US$8bn deal. "Tokens are the central currency for companies building with AI, and it's clear that the real-world economic potential will depend on making good use of scarce compute resources," says Patrick Collison, CEO of Stripe. "Stripe is building the economic infrastructure for AI and together with OpenRouter we'll help businesses maximise profitability by routing their requests intelligently and spending their tokens efficiently." While exact terms of the deal were not disclosed, the Financial Times reported that Stripe has agreed to buy OpenRouter for US$8bn, making it the Irish-American company's largest acquisition as it seeks to position itself at the centre of the AI economy. Hedging your bots. It may seem like an odd move that a financial services company is buying an API platform and marketplace that lets developers access hundreds of different AI models from various providers, such as OpenAI, Anthropic, Google and Meta - however it makes total sense when viewed through the lens of AI unit economics and token infrastructure. Stripe provides economic infrastructure for the world's fastest-growing businesses, including the vast majority of companies developing and building with AI. It already helps businesses maximise their revenue by optimising across a complex set of variables like payment methods, authorisation rates and fraud. OpenRouter has built a platform that helps businesses dynamically evaluate each request, routing it to the optimal model based on task complexity, price, speed and reliability, and is already used by the likes of NVIDIA, Zoom and Lovable. Stripe highlights that together with OpenRouter the firm will be able to help companies manage both sides of profitability - maximising revenue and efficacy while minimising costs. * The financial services company Stripe has announced that it has agreed to acquire OpenRouter * Stripe highlights that together with OpenRouter the firm will be able to help companies manage both sides of profitability * Stripe provides economic infrastructure for the world's fastest-growing businesses, including the vast majority of companies developing and building with AI * OpenRouter has built a platform that helps businesses dynamically evaluate each request, routing it to the optimal model based on task complexity, price, speed and reliability. A multi-model future. Many Chinese models such as those from Moonshot and Z.ai are quickly gaining ground among enterprise customers - narrowing the performance gaps while undercutting the top US firms on cost. According to The Information, OpenAI recently began giving OpenRouter a special discount to offer some OpenAI models at half of the price per token. Those models include GPT-5.6 Luna and GPT-5.6 Terra, which OpenAI originally released in June. The outlet reported that GPT-5.6 Sol, its most capable, flagship model, would also be discounted 50% on OpenRouter. At the moment, AI is not dominated by one firm alone; Anthropic and OpenAI are both reportedly gearing up for IPOs, while cheaper Chinese firms are gaining ground. On top of this, enterprises are increasingly asking staff to curtail usage - and in this environment, both cost and performance are emerging as key factors in model selection. OpenRouter's CEO Alex Atallah doesn't think one firm is going to come out on top from this environment. Discussing the Stripe deal, Alex says: "AI Magazine believe intelligence will be multi-model: no single model will be optimal for every task and developers need a neutral layer to orchestrate and manage them all. "Joining Stripe lets us accelerate that mission and bring the full AI ecosystem to every business." Company Portals