正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

This completely changes how I look at AI benchmarks. Qwen 3.8 Max Preview felt nearly unusable in OpenCode and Qwen Code. With the Claude Code harness, it suddenly came surprisingly close to Kimi K3 and Fable 5. Same model. Same task. Different harness. Wildly different result. How is this... show more

OmedTheVibeCoder

1,396 subscribers

30,223 次观看 • 3 个月前 •via X (Twitter)

科学技术游戏

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

You need to read this. Again. Qwen 3.8 Max Preview still performs badly in OpenCode and Qwen Code. Switch to the Claude Code harness, and it suddenly becomes dramatically better. This is reproducible. The harness difference is real.

You need to read this. Again. Qwen 3.8 Max Preview still performs badly in OpenCode and Qwen Code. Switch to the Claude Code harness, and it suddenly becomes dramatically better. This is reproducible. The harness difference is real.

OmedTheVibeCoder

18,655 次观看 • 3 个月前

Qwen 3.8 Preview is quietly catching Kimi K3 on frontend + game design.And this is just the preview. Full 3.8 is going to be a beast. We tested Qwen 3.8 Preview against Kimi K3 on 3D modeling and animation. Ran both through the same 3 prompts. Both on Max effort. Qwen 3.8 Preview was dramatically faster across the board, 3-4x quicker on every test. Kimi K3 needed 2 hours on the final scene; Qwen wrapped it in 31 minutes. Test 1 (Engine Sim) Qwen 3.8 Preview: 25min Kimi K3: 1hr 30min Test 2 (Voxel Dino Printer) Qwen 3.8 Preview: 8min Kimi K3: 45min Test 3 (Garden Scene) Qwen 3.8 Preview: 31min Kimi K3: 2hrs

Qwen 3.8 Preview is quietly catching Kimi K3 on frontend + game design.And this is just the preview. Full 3.8 is going to be a beast. We tested Qwen 3.8 Preview against Kimi K3 on 3D modeling and animation. Ran both through the same 3 prompts. Both on Max effort. Qwen 3.8 Preview was dramatically faster across the board, 3-4x quicker on every test. Kimi K3 needed 2 hours on the final scene; Qwen wrapped it in 31 minutes. Test 1 (Engine Sim) Qwen 3.8 Preview: 25min Kimi K3: 1hr 30min Test 2 (Voxel Dino Printer) Qwen 3.8 Preview: 8min Kimi K3: 45min Test 3 (Garden Scene) Qwen 3.8 Preview: 31min Kimi K3: 2hrs

GMI Cloud

25,513 次观看 • 3 个月前

I’m not even testing Opus 5 Max yet, guys. This is only High and xHigh—and the difference is already impossible to put into words. Kimi K3 and Qwen 3.8 Max Preview genuinely look like bots next to this. Opus 5 xHigh is on an entirely different level. Just LOOK at these details. This is the first model that has finally delivered exactly what I designed this benchmark to reveal.

I’m not even testing Opus 5 Max yet, guys. This is only High and xHigh—and the difference is already impossible to put into words. Kimi K3 and Qwen 3.8 Max Preview genuinely look like bots next to this. Opus 5 xHigh is on an entirely different level. Just LOOK at these details. This is the first model that has finally delivered exactly what I designed this benchmark to reveal.

OmedTheVibeCoder

31,464 次观看 • 3 个月前

Nah, this can’t be real. Same GPT-5.6 Sol. Same task. Same thinking level. I switched from regular Codex to the Claude Code harness, and the result instantly became way better. Claude Code is somehow making every model cook. GPT-5.6 SOL WAS ALREADY TOP TIER!! 😭

Nah, this can’t be real. Same GPT-5.6 Sol. Same task. Same thinking level. I switched from regular Codex to the Claude Code harness, and the result instantly became way better. Claude Code is somehow making every model cook. GPT-5.6 SOL WAS ALREADY TOP TIER!! 😭

OmedTheVibeCoder

127,040 次观看 • 3 个月前

GLM 5.3 vs Fable 5 vs Qwen 3.8 Max vs Grok 4.6 Same task. Who won?

GLM 5.3 vs Fable 5 vs Qwen 3.8 Max vs Grok 4.6 Same task. Who won?

Mario Nawfal

102,885 次观看 • 2 个月前

This is probably my most important post about DeepSeek V4 Flash. Do NOT judge this model after testing only one harness. Pi, Claude Code, and OpenCode produced dramatically different results—and OpenCode completely changed my verdict. You really need to read the details below.

This is probably my most important post about DeepSeek V4 Flash. Do NOT judge this model after testing only one harness. Pi, Claude Code, and OpenCode produced dramatically different results—and OpenCode completely changed my verdict. You really need to read the details below.

OmedTheVibeCoder

68,853 次观看 • 3 个月前

Alright guys, wow. Qwen 3.8 27B is basically Qwen 4. What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the sharks. Fable even had all of the fins reversed and details jumbled on the first turn. Before now, it was really a 2 to 8 turn test to get something final regardless of model. Qwen 3.8 27B just nailed this in one output, no harness at all, just purely the model thoroughly thinking before outputting the final result. On every asset used, different shark variants, everything, it all looks incredibly tight. The boat movement, shark speed and engagement, everything is so much better than I expected, and I was expecting a lot. Imagine what 8 turns would do in a harness? This confirms to me that while Qwen 3.8 27B thinks a lot, for the first time in a local model on the base weights, IT MIGHT ACTUALLY BE WORTH IT This is a Frontier Lab model result, and it ran on my 5090 at 60tps in Q5_K_M. I am just in complete astoundment at what Qwen has accomplished here, and I am incredibly grateful that this model is open source. The speed will also increase drastically with optimisations. This is just an incredible day. What a blessing. More to come!

Alright guys, wow. Qwen 3.8 27B is basically Qwen 4. What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the sharks. Fable even had all of the fins reversed and details jumbled on the first turn. Before now, it was really a 2 to 8 turn test to get something final regardless of model. Qwen 3.8 27B just nailed this in one output, no harness at all, just purely the model thoroughly thinking before outputting the final result. On every asset used, different shark variants, everything, it all looks incredibly tight. The boat movement, shark speed and engagement, everything is so much better than I expected, and I was expecting a lot. Imagine what 8 turns would do in a harness? This confirms to me that while Qwen 3.8 27B thinks a lot, for the first time in a local model on the base weights, IT MIGHT ACTUALLY BE WORTH IT This is a Frontier Lab model result, and it ran on my 5090 at 60tps in Q5_K_M. I am just in complete astoundment at what Qwen has accomplished here, and I am incredibly grateful that this model is open source. The speed will also increase drastically with optimisations. This is just an incredible day. What a blessing. More to come!

Kyle Hessling

26,243 次观看 • 2 个月前

I tested Ox Alpha vs Kimi K3 Ran Ox Alpha through the Claude Code harness. Honestly, I expected much more from Ox Alpha. IMO, it wasn't even close to GPT-5.6 Terra. Kimi K3 was the clear winner for me. What model could this be?

I tested Ox Alpha vs Kimi K3 Ran Ox Alpha through the Claude Code harness. Honestly, I expected much more from Ox Alpha. IMO, it wasn't even close to GPT-5.6 Terra. Kimi K3 was the clear winner for me. What model could this be?

Bhavy☄️

106,084 次观看 • 2 个月前

I’ve tried driving Qwen 3.6 on my MacBook Pro with a few different agent harnesses: Claude Code: too slow. It moves like a turtle. Codex: Qwen stopped in the middle of tasks, probably due to tool-call mismatches between them. Qwen Code: so far the best experience. Very smooth. Qwen Code sometimes even tells a joke while you’re waiting. This video is played at original speed so you can get a real feel for how Qwen works.

I’ve tried driving Qwen 3.6 on my MacBook Pro with a few different agent harnesses: Claude Code: too slow. It moves like a turtle. Codex: Qwen stopped in the middle of tasks, probably due to tool-call mismatches between them. Qwen Code: so far the best experience. Very smooth. Qwen Code sometimes even tells a joke while you’re waiting. This video is played at original speed so you can get a real feel for how Qwen works.

Michael Guo

20,577 次观看 • 5 个月前

Claude Fable 5 vs Qwen3.8-Max i build Mario game with the same single prompt in both models qwen was 7.5x cheaper than fable yet it competes with one of the strongest AI model is China finally leading the AI race in open models?

Claude Fable 5 vs Qwen3.8-Max i build Mario game with the same single prompt in both models qwen was 7.5x cheaper than fable yet it competes with one of the strongest AI model is China finally leading the AI race in open models?

Tim Jayas

10,433 次观看 • 3 个月前

How to run GLM-5.2 (Or Minimax or Qwen or Deepseek or Kimi) in Claude Code No JSON editing. No vim 😮‍💨 1. Install Claude Code npm install -g Paul Jankura-ai/claude-code 2. Grab a Fireworks API key → 3. Install FireConnect + drop in your key (one line): curl -fsSL | FIREWORKS_API_KEY="fw_..." bash 4. Flip it on: fireconnect claude on 5. Switch to GLM-5.2 and bam! Swap models with: fireconnect claude model select Docs to do the same for Pi, Opencode, and Codex:

How to run GLM-5.2 (Or Minimax or Qwen or Deepseek or Kimi) in Claude Code No JSON editing. No vim 😮‍💨 1. Install Claude Code npm install -g Paul Jankura-ai/claude-code 2. Grab a Fireworks API key → 3. Install FireConnect + drop in your key (one line): curl -fsSL | FIREWORKS_API_KEY="fw_..." bash 4. Flip it on: fireconnect claude on 5. Switch to GLM-5.2 and bam! Swap models with: fireconnect claude model select Docs to do the same for Pi, Opencode, and Codex:

Sinan Ozdemir

149,913 次观看 • 4 个月前

Hit your usage limit in Codex/Claude Code mid-task? Orca now lets you resume the exact same session in a different agent harness — one click. Codex ⇄ Claude Code, full context intact🔥

Hit your usage limit in Codex/Claude Code mid-task? Orca now lets you resume the exact same session in a different agent harness — one click. Codex ⇄ Claude Code, full context intact🔥

Orca ADE

55,781 次观看 • 3 个月前

🔥 Running Kimi K3 (Kimi.ai) inside Claude Code is a game-changer. Claude is the best coding harness out there, and now you can pair it with Kimi's powerful (and affordable) K3 model running 24/7 on a VPS. Loops, subagents, and max performance. Watch the full setup guide 👇

🔥 Running Kimi K3 (Kimi.ai) inside Claude Code is a game-changer. Claude is the best coding harness out there, and now you can pair it with Kimi's powerful (and affordable) K3 model running 24/7 on a VPS. Loops, subagents, and max performance. Watch the full setup guide 👇

Boxmining

70,682 次观看 • 3 个月前

Claude Fable 5 vs Kimi K3 — same 3D build, 3.3x cheaper. someone prototyped a 3D stadium seat-view picker with Fable 5 — preview exactly what you'll see from your seat before buying a ticket. I rebuilt the same thing with Kimi K3 in two hours, at 3.3x lower cost. no more static seat maps. pick a seat, see the actual view in 3D, then buy. works for cinemas, concerts, cricket grounds, Wimbledon — any venue. K3 dropped this week: 2.8 trillion parameters, open weights July 27. 3D is the future, and it just got a lot cheaper to build. Live Fable 5: Live Kimi K3: Code: full Kimi K3 breakdown below.

Claude Fable 5 vs Kimi K3 — same 3D build, 3.3x cheaper. someone prototyped a 3D stadium seat-view picker with Fable 5 — preview exactly what you'll see from your seat before buying a ticket. I rebuilt the same thing with Kimi K3 in two hours, at 3.3x lower cost. no more static seat maps. pick a seat, see the actual view in 3D, then buy. works for cinemas, concerts, cricket grounds, Wimbledon — any venue. K3 dropped this week: 2.8 trillion parameters, open weights July 27. 3D is the future, and it just got a lot cheaper to build. Live Fable 5: Live Kimi K3: Code: full Kimi K3 breakdown below.

Kirill

207,659 次观看 • 3 个月前

Harness Engineering 2.0 is here! We benchmark models all the time but the harness running the model can dramatically improve the result, cost, and execution time too. And finally, devs just open-sourced Harness Arena, where you can test the same task across multiple harness and model combinations side by side. You provide one task and then run it across several model x harness configs that you want to test. Every config receives the same prompt, files, tools, skills, and evaluation checks. Harness Arena then returns: - total cost - execution time - whether the task passed - files and artifacts produced If you want to try it yourself, it is implemented within the open-source HarnessRouter project. Here's the repo: (don't forget to star it ⭐ ) HarnessRouter is a plug-and-play infrastructure layer that lets you run any harness under a single interface, like: - Codex - Hermes - Claude code - DeepSeek Harness - System One, powered by Jev - And 25 more agent harnesses In Harness Arena, the cheapest run does not automatically win. A config must complete the task correctly first. Cost and execution time are compared only among the configurations that pass. Here's the result from one Harness Arena test: - Hermes + GPT-5.2: 5/5 successful runs using 0.47 credits - Claude Code + Opus 4.8: 3/5 successful runs using 223 credits That is around 474x difference in credits, while the more expensive configuration also failed two runs. You can keep the model fixed to measure how much the harness changes the result. Or vary both to find the best harness and model pairing for that task. This helps answer practical questions such as: - Should this task run on Codex or Claude? - Can a cheaper config still pass every check? - Which setup is best to handle this workload? Once you find the best config, you can run future tasks through it using the same Unified Harness Protocol (UHP) interface for sessions, streaming, files, artifacts, cancellation, and failures. So you are no longer choosing an agent stack based only on a model leaderboard. Instead, you can benchmark the complete runtime on the task your product actually needs to solve. To dive deeper, my co-founder wrote an article explaining why model routing is different from harness routing, how UHP works, and what supporting multiple agent runtimes involves. It also covers the local setup, a working request, and how sessions and files move through the system. Read it below.

Harness Engineering 2.0 is here! We benchmark models all the time but the harness running the model can dramatically improve the result, cost, and execution time too. And finally, devs just open-sourced Harness Arena, where you can test the same task across multiple harness and model combinations side by side. You provide one task and then run it across several model x harness configs that you want to test. Every config receives the same prompt, files, tools, skills, and evaluation checks. Harness Arena then returns: - total cost - execution time - whether the task passed - files and artifacts produced If you want to try it yourself, it is implemented within the open-source HarnessRouter project. Here's the repo: (don't forget to star it ⭐ ) HarnessRouter is a plug-and-play infrastructure layer that lets you run any harness under a single interface, like: - Codex - Hermes - Claude code - DeepSeek Harness - System One, powered by Jev - And 25 more agent harnesses In Harness Arena, the cheapest run does not automatically win. A config must complete the task correctly first. Cost and execution time are compared only among the configurations that pass. Here's the result from one Harness Arena test: - Hermes + GPT-5.2: 5/5 successful runs using 0.47 credits - Claude Code + Opus 4.8: 3/5 successful runs using 223 credits That is around 474x difference in credits, while the more expensive configuration also failed two runs. You can keep the model fixed to measure how much the harness changes the result. Or vary both to find the best harness and model pairing for that task. This helps answer practical questions such as: - Should this task run on Codex or Claude? - Can a cheaper config still pass every check? - Which setup is best to handle this workload? Once you find the best config, you can run future tasks through it using the same Unified Harness Protocol (UHP) interface for sessions, streaming, files, artifacts, cancellation, and failures. So you are no longer choosing an agent stack based only on a model leaderboard. Instead, you can benchmark the complete runtime on the task your product actually needs to solve. To dive deeper, my co-founder wrote an article explaining why model routing is different from harness routing, how UHP works, and what supporting multiple agent runtimes involves. It also covers the local setup, a working request, and how sessions and files move through the system. Read it below.

Avi Chawla

30,196 次观看 • 2 天前

I gave the exact same prompt to Qwen 3.7 Max and GLM 5.2. Same workflow. Same platform. Nearly the same starting price. The results weren't even close. Here's what happened. 👇

I gave the exact same prompt to Qwen 3.7 Max and GLM 5.2. Same workflow. Same platform. Nearly the same starting price. The results weren't even close. Here's what happened. 👇

Md Riyazuddin

153,142 次观看 • 3 个月前

Here's Codex talking to Fable 5 w/ my plugin. Unlike calling the API, `claude-channel-cli` talks to the same full Claude Code harness as you. This means you can talk to either AI & they'll guide each other naturally with shared memories, decisions, and research.

Here's Codex talking to Fable 5 w/ my plugin. Unlike calling the API, `claude-channel-cli` talks to the same full Claude Code harness as you. This means you can talk to either AI & they'll guide each other naturally with shared memories, decisions, and research.

Aaron Ng

22,145 次观看 • 4 个月前

Kimi K3 vs Fable 5. Same prompt and reference. 3D globe dashboard. Kimi left, Fable right. To me, Kimi's UI fidelity is a touch better. The globe looks better too, with richer textures and smoother motion. This task: ~$3 with Kimi, ~$15 with Fable. I'd pick Kimi k3 for frontend work.

Kimi K3 vs Fable 5. Same prompt and reference. 3D globe dashboard. Kimi left, Fable right. To me, Kimi's UI fidelity is a touch better. The globe looks better too, with richer textures and smoother motion. This task: ~$3 with Kimi, ~$15 with Fable. I'd pick Kimi k3 for frontend work.

Kai

227,135 次观看 • 3 个月前

Qwen 3.8 will be benchmaxxed slop second to fable 5 is such a hype statement , its not even close to kimi k3 and GPT-5.6 sol you can test it yourself on the qwen chat website currently available > token hungry > not that good as advertised , will test more The best part of kimi was that in blog they themselves wrote its not comparable to fable 5 and sol 5.6 in bunch of user experience stuff We need more truth like this in our industry , other than that i will run my vision test

Qwen 3.8 will be benchmaxxed slop second to fable 5 is such a hype statement , its not even close to kimi k3 and GPT-5.6 sol you can test it yourself on the qwen chat website currently available > token hungry > not that good as advertised , will test more The best part of kimi was that in blog they themselves wrote its not comparable to fable 5 and sol 5.6 in bunch of user experience stuff We need more truth like this in our industry , other than that i will run my vision test

Chetaslua

127,047 次观看 • 3 个月前

Someone built a plugin that cuts Claude Code's cost by 54% on the same task, same model. Not a different model. Not a different harness. Just smarter tools underneath. It's called WOZCODE and the benchmark numbers are genuinely CRAZYYYY.

Someone built a plugin that cuts Claude Code's cost by 54% on the same task, same model. Not a different model. Not a different harness. Just smarter tools underneath. It's called WOZCODE and the benchmark numbers are genuinely CRAZYYYY.

Hasan Toor

20,754 次观看 • 6 个月前