Claudiu Micu | Building Wakebound @_MrCode_

DeepSeek V4.1 Flash 1017 scores 45 on the Artificial Analysis Intelligence Index, improving on DeepSeek V4.1 Flash by 6 points, and is the equal #2 open-weights model @deepseek_ai has released DeepSeek V4.1 Flash 1017, an update on their flagship model DeepSeek V4.1 Flash. On the first-party API, DeepSeek V4.1 Flash 1017 is priced at $0.30 per 1M input tokens and $1.20 per 1M output tokens during peak hours, with off-peak pricing half of peak rates, the same pricing as DeepSeek V4.1 Flash. Key results: ➤ DeepSeek V4.1 Flash 1017 makes gains in agentic capabilities, knowledge and scientific reasoning. DeepSeek V4.1 Flash 1017’s Elo rating on GDPval-AA v2.1 is 1746, an increase of 146 points from DeepSeek V4.1 Flash’s rating of 1600. On AA-Briefcase v1.1, its Elo rating increases by 150 points from 1420 to 1570. On AA-Omniscience, its non-hallucination rate is 33%, a 29-point improvement from DeepSeek V4.1 Flash’s 4%, while seeing an improvement in accuracy from 43% to 46%, pushing its overall AA-Omniscience Index score from -5 to +5. On CritPt, DeepSeek V4.1 Flash 1017 scores 24%, up from 14% for DeepSeek V4.1 Flash. ➤ DeepSeek V4.1 Flash 1017 is now the #2 open-weights model alongside GLM-5.3. With a score of 45 on the Artificial Analysis Intelligence Index, DeepSeek V4.1 Flash 1017 is behind MiMo-V2.6-Pro by only 1 point and ahead of other open-source models like Step 5 Preview and Kimi K3 (both at 44). ➤ With a Cost per Task of $0.26, DeepSeek V4.1 Flash 1017 sits in the most attractive quadrant of Capability vs. Cost. With a Cost per Task of $0.26 and Intelligence Index score of 45, DeepSeek V4.1 Flash 1017 compares favorably to GLM-5.3-Flash at $0.28 per task and a score of 42. At the same time, MiMo-V2.6-Pro (46) at $0.13 per task and GPT-6.1 Sol (medium, 48) at $0.21 per task both offer higher capability at a lower Cost per Task. Additional model details: ➤ Context window: 1M tokens (equivalent to DeepSeek V4.1 Flash) ➤ Pricing: $0.30 per 1M input tokens and $1.20 per 1M output tokens, unchanged from DeepSeek V4.1 Flash. Cache hit price of $0.006 per 1M tokens, a 98% discount ➤ Size: 552B total parameters (16B active) ➤ License: MIT ➤ Providers: Available through DeepSeek’s first-party API, @FireworksAI_HQ , @baseten, @DeepInfra , @wafer_ai , @Cloudflare and others

Who's winning compute? Biggest AI site Epoch AI tracks right now: xAI's Colossus 2, ~946 MW. Anthropic + Amazon's New Carlisle: ~910 MW. Meta's Prometheus: ~471 MW. xAI and Anthropic: same power class, ~20x apart on revenue ($3.3B vs $67.2B). Watts aren't the moat.

4 of the top 10 labs on the scoreboard are Chinese: Qwen #5, Kimi #7, Z.ai #8, DeepSeek #9. Fun one: on @arena, Kimi K3 (1487) ranks above OpenAI's best, GPT-5.6 Sol (1485). Moonshot is valued at $50B. OpenAI at $852B. 17x the price for 2 fewer Arena points.

@CloudDogx @arena Fair read. The rest of Mistral's numbers point the same way: $24B valuation, #111 on Arena, #50 on Artificial Analysis. EU buyers are paying for hosting and data control, not leaderboard spots. The open question is how long that premium holds if the gap keeps growing.

Europe's AI champion, by the numbers: Mistral: $24B valuation (Reuters, Sept). Best @arena rank: #111 (Mistral Large 4, 1432). The leader, Gemini 4 Argon: 1524. 92 points is a lot of ground. Writing this from Romania, I really want them to close it.

Monday standings, 19 Oct: 1. Anthropic 94.4 2. Google DeepMind 92.3 3. OpenAI 66.9 Biggest riser: DeepSeek, 41.2 → 42.6. Its new V4.1 Flash 1017 lifted its Artificial Analysis index from 39.5 to 45.1. Full top 10: airacescoreboard.com/?utm_source=x&…

Several things he said on this pod: >OpenAI and Anthropic reaching $600–700bn combined revenue by December 2027. >Anthropic revenue growing roughly 10x this year, from below $10bn to above $100bn (run-rate revs) >Anthropic potentially reaching ~$1tn revenue in 2028–29. He says it could become the first $10tn company >An illustrative $2tn Anthropic IPO valuation at 20x $100bn revenue. He also discusses a $20tn valuation scenario with several trillion dollars of revenue and lower multiples. >Anthropic inference gross margins around 75% or higher. He says revenue already covers training and inference compute costs, with OpenAI reaching that point in another 2–3 quarters. >OpenAI and Anthropic each attempting to add 5–10 GW of compute capacity in 2027, versus a few GW at Google. >Roughly $2tn of US capex next year across AI infrastructure and its supply chain. >His payback example: $1tn invested, $1.5tn cumulative revenue over six years, or $250bn annually; six-year GPU lives and 15-year data-center lives.

Fireside Alpha @firesidealpha

a week ago

SemiAnalysis's Dylan Patel predicts OpenAI and Anthropic could reach $600–700 billion in combined revenue by December 2027--and raises the possibility of $1.5 trillion by year-end “As far as AI’s impact on the U.S. economy, if AI revenue is only $3 trillion in 2031, then, yeah,

47 35 386 371K 351

@finnsdesk Agree, daily use is the eval that matters most. The boards are split too: Claude Opus 5.5 is #1 on Artificial Analysis, while Arena has Claude at #3 and OpenAI's best down at #21. What does Codex win at for you that Claude doesn't?

Benchmarks can't agree on who's best: Gemini 4 Argon: #1 on Arena, #9 on Artificial Analysis. Claude Opus 5.5: #1 on AA, #3 on Arena. Meta's Muse Wonder: #2 on Arena, 1 point ahead of Claude. Vibes vs evals. Pick your leaderboard, pick your winner.

@Nitocnt Neither on its own, honestly. Arena is people voting on answers they like, so it rewards tone and polish. AA is a fixed eval suite, closer to raw capability. That's why I blend both with revenue, valuation and Polymarket odds instead of crowning one board.

Exciting news: DeepSeek-V4.1-Flash-20261017 (Max) by @deepseek_ai is now the #1 open model in Code Arena: WebDev, surpassing Kimi K3 (Max) at 10x lower cost! Highlights: - #1 open model, ahead of Kimi K3 (Max) - #12 overall with 1670 pts, ahead of Claude Opus 5 (High) and within ~10 pts of GPT-6 Sol (Max) - Shifts the Pareto frontier at $0.3/$1.2 per million input/output tokens - +52 pts vs. the previous DeepSeek-V4.1-Flash-20260910 Congrats to the @deepseek_ai team on another strong contribution to the open source ecosystem!

keyboard_arrow_left Previous keyboard_arrow_right Next

DeepSeek @deepseek_ai

3 days ago

🚀 DeepSeek-V4.1-Flash-1017 is here! Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API. 🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy,

keyboard_arrow_left Previous keyboard_arrow_right Next

2K 4K 37K 7.3M 7K

OpenAI, money vs models: Valuation: $852B last round, now in talks at $1.4T (Bloomberg). Best @arena rank: #21. Best Artificial Analysis rank: #8. Polymarket odds of best model by end of 2026: 7.3%. Investors and benchmarks are scoring two very different OpenAIs.

@Nitocnt Depends on the job. Arena is real people blind-voting, so it measures what feels best to use. AA is hard evals, so it measures reasoning and code. For chat I'd trust Arena, for building AA. I track both ranks for every lab here: airacescoreboard.com

We’re committing a portion of Anthropic’s shares to the Anthropic Public Benefit Fund, a new wholly owned entity dedicated entirely to the public benefit. It will start off with 5% of our shares. For every 50% increase in the company’s value over a set baseline, we can contribute up to an extra 2%, subject to our board’s discretion. In total, this means that roughly 20% of Anthropic’s shares are committed to the public good. Our co-founders have already pledged to donate 80% of their equity, and around half our employees are donating a portion of their equity as part of a donation matching program we’ve run since our founding. In the post below, we share more details about our new fund. anthropic.com/news/public-be…

@amigofriend75 Same settlement source for both: the Arena text leaderboard (style control off), checked on the end date. So it's mostly the window. Right now Gemini 4 Argon leads Claude Opus 5.5 by 16 points (1524 vs 1508). The Dec market is betting Anthropic closes that before Dec 31.

Polymarket, on who has the best AI model: End of October: Google, about 3 in 4. End of 2026: Anthropic 47.5% vs Google 39.5%. Same race, opposite favorite, two months apart. I put the odds next to the model rankings and the money: airacescoreboard.com/?utm_source=x&…

🚀 DeepSeek-V4.1-Flash-1017 is here! Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API. 🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model. 1/4

keyboard_arrow_left Previous keyboard_arrow_right Next

Arena Alignment Index update: Two new labs and three new models have landed! - Meta’s new top model is Muse Wonder, replacing Muse Spark 1.3. - Reflection debuts with Beam 1.0, the #1 US open model on Agent Arena overall. - StepFun, another Chinese lab, joins the leaderboard with Step 5 Preview. Arena Alignment Index rankings by lab: - #6 @AIatMeta: Muse Wonder (Max) at 77.5% - #7 @reflection_ai: Beam-1.0-501B-A23B at 75.3% - #8 @StepFun_ai: Step 5 Preview at 75.1% Among the three new model entrants: - Meta is strongest with the highest Alignment Index score (77.5%). It also has the lowest unauthorized action rate (2.09%) and deceptive completion rate (14.41%) of the three. - Reflection AI is strongest at avoiding false attribution: 4.50%. - StepFun does not lead a category, but its strongest relative result is unauthorized action at 2.11%, almost tied with Muse Wonder’s 2.09%.

Arena.ai @arena

2 weeks ago

Introducing the Arena Alignment Index, our new benchmark measuring safety and alignment of AI agents in real-world use. Built from 90K+ real-world agent sessions across 27 models, the index measures three critical signals: - Unauthorized Action (UA): Taking actions beyond the

44 55 610 310K 102

xAI (@spacexai): ~$250B implied value in the SpaceX merger, on ~$3.3B annualized revenue. And that revenue counts X ads and subscriptions. That's ~76x revenue. Anthropic and OpenAI sit at roughly 15–17x. Somebody's paying for the future here, not the P&L.

airacescoreboard.com