正在加载视频...

视频加载失败

Same model. Up to 65% less. DeepSeek’s new API prices take effect on August 16. Marathon lets latency-tolerant workloads trade wait time for lower inference costs without switching models. ▷ Choose NOW when every second matters. ▷ Choose SOON or LATER when a few minutes are acceptable. ▷ Choose...

15,113 次观看 • 1 个月前 •via X (Twitter)

2 条评论

AI BIRD 的头像
AI BIRD1 个月前

Hi Marathon Team, I’ve been using Marathon Delayed Inference and found that tasks are constantly stuck in queued state. How should I fix this? I can supply more details if required, thanks.

BlockWeb 的头像
BlockWeb1 个月前

The ability to trade latency for cost is a pretty smart model.

相关视频

There is no best model. There's a lot of noise about models right now. Who is training them, who owns them, where legal intelligence should live. One question actually matters: what produces the best outcome for the legal task in front of you? That's how we decide things at Legora. We optimize for the end-to-end outcome on a legal task. The model is one layer of that system, not the system. Models are uneven and the frontier changes almost weekly. One model plans a long job well, another runs deep analysis across thousands of documents. Some have to be told exactly what to do, and some are fine with a vague brief. They all break in different ways. So our lawyers write evals and we test them with the Legora BAR, our benchmark for agentic reasoning. Every model takes every test, and the model that wins gets the work. We post-train when we know it buys our customers better performance on a specialized task. Training is a tool we reach for when it helps, nothing more than that. The intelligence that compounds sits in the orchestration layer. Precedents, review standards, client requirements. That knowledge has to stay editable, auditable and portable. In our system, a changed review standard is an edit that takes effect the same day, with no new model training required. No lawyer should have to worry about which model did the work, any more than they think about which chip is in their laptop. They should only care about the quality of the work. That's what we are focused on. If you want the engineering version of this argument rather than the CEO version, our CPO, Bryan Tsao, and CTO, Jacob Lauritzen, take it apart in the video below.

Max Junestrand

49,101 次观看 • 10 天前

your agent has thirty tools. it calls two of them. the other twenty eight are not sitting idle somewhere. they are in the request, every request, and they are doing damage in two places at once. first the obvious one. tool schemas go into the prompt, and a schema is not a name. it is a description, a parameter list, types, required fields, an example. thirty of those is a few thousand tokens that ship with every single call, including the ones where the agent just says thanks and stops. you are paying rent on twenty eight tools that have never fired. second, and this is the one that costs more. when the request says cancel the order, the model picks by matching against everything available. four of your tools are plausible: cancel_order, refund_order, update_order, void_order. it is choosing among them based on the descriptions you wrote, one afternoon, months ago. every tool you add is another candidate in that shortlist. the twenty eight you never call are not neutral. they are noise in the one decision that determines whether the run works. > why it grows without anyone deciding to nobody adds thirty tools on purpose. you add one for a task, it works, it stays. six months later the registry is a catalogue and no one has ever removed anything, because removing a tool feels risky and adding one feels free. and there is no feedback telling you otherwise. the unused ones never error. they never appear in a failing trace. they are invisible in exactly the way that lets them accumulate. > what to actually do count calls per tool over the last thousand runs. this is one group-by and it usually shocks people. the ones at zero are pure cost. ship the tools the task needs, not the whole registry. a research phase does not need deploy. a writing phase does not need the database. swap the set between phases instead of loading everything up front. same agent, different tools, depending on where the run is. and when two tools could both plausibly answer the same request, that is not redundancy you can ignore. it is a coin flip you built into the system. the twenty eight tools are not unused. they are used every time, by the part of the run you cannot see.

Hanako

24,656 次观看 • 1 个月前

THIS WALLET STACKED $230K ON BTC UP/DOWN BETS. THE BLUEPRINT TO AUTOMATE THE SAME EDGE WITH CLAUDE The wallet is $230K all-time, every position a Bitcoin or Ethereum Up or Down market It never guesses direction. It enters only when the math and the market disagree THE STRATEGY: BTC moves are not fully random. When the market enters a committed directional state, continuation is measurable. That is Markov persistence Entry signal: > Δ = p̂ − q ≥ ε Model probability minus market price. Enter only on a 5% gap or more Persistence filter: > p(j*,j*) ≥ 0.87 Only trade states with 0.87 persistence or higher. Below that, skip. This is what holds the win rate above 65% with zero directional guessing Payout: > r = (1 − q) / q At q = 0.647 that is +54.5% a win. At q = 0.441, +126.7%. Lower entry price, bigger asymmetry Sizing: > f* = p − (1−p)/b Kelly. At p = 0.87, b = 0.647, f* ≈ 0.71. Size to the edge, never to gut HOW TO BUILD IT WITH CLAUDE: What separates this from a static bot: Claude reads its own trade journal every night and rewrites its own thresholds 1. Take an open-source Polymarket bot repo as your base logic. Feed it to Claude and have it migrate to CLOB v2: py_clob_client_v2, Safe wallet support, fee-aware evaluation 2. Hard-code the filters. Enter only when Δ ≥ 0.05 and p(j*,j*) ≥ 0.87. Apply Kelly on every fill. 3. Run DRY_RUN first. Log every signal, entry price, Markov state, and simulated P/L. No real money until the numbers hold for days 4. The nightly loop. Claude reads the journal, finds which persistence states actually won, adjusts MIN_PROB and MIN_EDGE, ships tomorrow's rules. The agent is sharper after 50 to 100 trades THE SETUP: Claude Opus as the brain. An open-source repo as the starting logic. A Polygon wallet with $50 to $100. Telegram for the morning report Start at $1 to $2 per trade while it learns. Scale only when the dry runs and the live fills line up 17,000 trades compound a thin edge into six figures. The model finds the edge. The nightly loop keeps it sharp Bookmark before you point a bot at your first window

Yarchi

22,966 次观看 • 4 个月前

Production traffic is not uniform. You get a few requests that need your best model, but most are simple questions and lookups you can solve with cheaper, faster models. The most expensive mistake you can make today is sending every request to your strongest model. You need routing. Period. This is the simplest trick to improve the architecture of whatever you are building. Please, don't implement routing yourself. You don't have to. I'm currently working with TrueFoundry's Auto Routing. It reads each request, classifies it as simple, medium, or complex, and sends it to the model assigned to that tier. You have two choices: 1. Send every request to the free heuristic classifier to score signals such as technical vocabulary, code, prompt length, and multi-step reasoning. 2. Send the request to an LLM classifier when its difficulty requires a more nuanced judgment. The beauty of using routing is that nothing changes in your code. You still call a single endpoint model, but routing works behind the scenes to pair every request with the best possible model. TrueFoundry ran several experiments with two different setups: 1. Send every request to Claude Opus 2. Send every request to a router with Haiku, Sonnet, and Opus The first experiment ran 550 deterministically graded academic prompts through every setup. Auto Routing was 69% cheaper while retaining 98% of the baseline quality. The second experiment ran three production-shaped workloads through every setup, using user chats, developer chats, and long agent tasks. Auto Routing was 80% cheaper. Thanks to the TrueFoundry team for partnering with me on this post.

Santiago

15,543 次观看 • 16 天前