Obolith @Obolith42
AI infrastructure pricing, cost tools & GPU economics. Find the cheapest way to run AI — or what your GPU could earn. obolith.com Joined March 2024-
Tweets53
-
Followers9
-
Following137
-
Likes19
What would routing save on your workload? Run your own traffic mix and model prices through the calculator: obolith.com/calculators?ut…
$6.00 → $0.60 per 1M tokens: a routing scenario, not a guaranteed saving. Send the easy 80% of requests to a small model and the blended cost drops 10×. Your result depends on your traffic mix and both models' prices.
Prices are collected daily. Current spot rates: H100 SXM spot: obolith.com/gpu/h100?utm_s… B200 SXM spot: obolith.com/gpu/b200
+21% in 30 days. Verda's H100 SXM and B200 SXM spot rates both rose. H100 SXM spot: $1.63/h → $1.96/h B200 SXM spot: $3.06/h → $3.71/h (2026-09-10 → 2026-10-10) Budget on last month's spot rate and you're already off.
Side-by-side API prices, checked daily: obolith.com/models/claude-…
Same $0.10 input price. Haiku 5.5 output: $0.50 vs Scout's $0.30 per 1M tokens. Claude Haiku 5.5: $0.10 in / $0.50 out, 1M context. Llama 4 Scout: $0.10 / $0.30, 1.3M context. Qwen3.7 Flash: $0.03 / $0.13, 1M context. Price only, no quality claim. As of 2026-10-10.
Exactly! The interesting question is how much further costs can drop with open models while maintaining the same classification accuracy. Do you have a concrete pricing example for this kind of workload on Batchrate? We’d love to compare it on Obolith.
If a job doesn't need an answer in seconds, why pay full price per token? OpenAI, Anthropic and Google all run a batch API at 50% off. The trade: no low-latency response. Completion is usually within 24 hours, often much sooner.
Example workload: 2M classification calls a month, 600 in + 20 out tokens each, on the batch API (−50%). Full breakdown in the guide: obolith.com/guides/batch-a…
5/5 So recheck your model. When did you last check your model's output rate? Output price per 1M tokens for each model, collected daily: obolith.com/llm-api-pricin…
4/5 The counterpoint: Qwen3 14B output went $0.91 → $0.22 per 1M tokens (-76%). Prices move both ways. Even inside one family, the direction differs by model.
1/5 GLM 5.2 output pricing rose 131% between 2026-09-10 and 2026-10-10. $3.04 → $7.00 per 1M output tokens. LLM API prices don't only go down.
🚀 Obolith launches on DevHunt on October 20 Compare AI infrastructure costs in one place: 🤖 LLM API pricing ⚡ Cloud GPU rentals 📊 Vector databases & more We're building a free tool to help developers spend less on AI Support our launch! 🙌 devhunt.org/tool/obolith #DevTools
A40 on-demand prices by provider, as of 2026-10-10: obolith.com/gpu/a40?utm_so…
+20% in 7 days: RunPod A40 and L4 on-demand went from $0.49/h to $0.59/h (2026-10-03 → 2026-10-10). For the A40, Vast.ai (marketplace) is at $0.469/h. How often do you recheck your GPU hourly rate?
4/4 B200 192GB: Hyperstack $6.00/h to RunPod $7.99/h. That's 1.3×. Cheapest tracked provider vs priciest: 25% less on B200, 96% less on V100. What would justify paying more for the same GPU? Full provider-spread breakdown: obolith.com/guides/same-gp…
3/4 Large spreads aren't limited to marketplace listings. RTX A4000 16GB: Hyperstack $0.15/h, RunPod $0.25/h, Paperspace $0.76/h. RTX A5000 24GB: RunPod $0.27/h, Paperspace $1.38/h. No Vast.ai in either. Both: 5.1× between cheapest and priciest.
1/4 V100: 28× between the cheapest and priciest tracked providers. B200: 1.3×. Same GPU, same day, very different invoice. 1 GPU on-demand, prices as of 2026-10-10.
kk @summerlilw
2 Followers 225 Following
Andy Long @longandy
1K Followers 4K Following AI, datacentres, telecoms, infosec, cycling, red sox baseball & Bitcoin. Tec diver. Ironman.
Batchrate @batchrate
11 Followers 50 Following LLM inference for jobs that can wait. OpenAI-compatible Batch API, results within hours. Big LLM jobs. Small bills.
Choums @7Chams
25 Followers 176 Following Si tu gagnes c’est normal, si tu perds y a r y aura remontada
MTz ST @mtz_st
67 Followers 697 Following
Nancy E @brahim66832982
17 Followers 1K Following Small girl, endless dreams, sometimes lonely heart 💫
Crecer con tus finanz... @joynul33
117 Followers 2K Following BOLSA 📈 FINANZAS💰AFILIADOS 🧑🧑🧒🧒DESARROLLO PERSONAL 💡📊 ¿OBJETIVO? LLEGAR A LOS 100.000€ 🚀🧨.。。。
Andy Long @longandy
1K Followers 4K Following AI, datacentres, telecoms, infosec, cycling, red sox baseball & Bitcoin. Tec diver. Ironman.
Tyler X @Tylertruthfx
2K Followers 3K Following Buddhist Wisdom | 大乘佛法 | ElvarOne Agent | ElvarOne Radio FM | AgoraDM Agent 通讯 | https://t.co/RaGfdtLml5 | 禅 | Forward Deploy Engineer |
Rubel | BooklierAi.co... @builtwithrubel
5K Followers 6K Following 📚https://t.co/PHl93kHtwi $120/m 🕵️https://t.co/k82pqzctch 0/m 💡https://t.co/825QFzNvOT 0/m
Abdullah Sajid @abdullahsajid_x
84 Followers 84 Following Chief Product Officer & Co-Founder, @hoppflow
Batchrate @batchrate
11 Followers 50 Following LLM inference for jobs that can wait. OpenAI-compatible Batch API, results within hours. Big LLM jobs. Small bills.
Emad @EMostaque
348K Followers 534 Following Building open source AI we own Founder @ii_posts, @StabilityAI Consistent inference is possible.
Gradio @Gradio
58K Followers 12 Following Build and share machine learning apps in 3 lines of Python. Part of the @Huggingface family 🤗.
Aran Komatsuzaki @arankomatsuzaki
198K Followers 446 Following Sharing AI research. Early work on AI (GPT-J, scaling, MoE). Ex ML PhD (GT) & Google.
CORSAIR @CORSAIR
2.3M Followers 619 Following The leading source for PC components and gaming hardware since '94. Need support? Send a DM to @CorsairSupport or visit: https://t.co/6JmSweDqDb
Anthropic @AnthropicAI
1.8M Followers 2 Following We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant @claudeai on https://t.co/FhDI3KQh0n.
Stability AI @StabilityAI
264K Followers 10 Following We’ll help you make it like nobody’s business. Multimodal media generation and editing tools to get your idea to production. Self-deploy? 👍 Need a partner? 🤝
LangChain @LangChain
272K Followers 206 Following Own your intelligence. Makers of LangSmith, @LangChain_OSS, and @LangChain_JS.
LlamaIndex 🦙 @llama_index
122K Followers 39 Following Turn any doc into AI-ready context. The AI research lab for document parsing and extraction. Developer-first APIs. LlamaParse: https://t.co/yQGTiRSfFL
Choums @7Chams
25 Followers 176 Following Si tu gagnes c’est normal, si tu perds y a r y aura remontada
Google Gemma @googlegemma
109K Followers 0 Following The official home of Google's Gemma. Lightweight, state-of-the-art open models by Google DeepMind, built on Gemini tech. What will you build? 🚀💻
Arena.ai @arena
237K Followers 224 Following Where AI meets the real world. We measure and advance the frontier of AI through community-driven evaluation. We’re hiring → https://t.co/XBZCrsdD77
Codex Releases @CodexReleases
28K Followers 3 Following Unofficial Codex release updates creds @OpenAI
Melvin Vivas @melvindvivas
8K Followers 764 Following AI Coding, Agents, Local Models AI devt https://t.co/MtiGBNIYAq youtube https://t.co/HRuCdgM6bD free Grok Bot templates https://t.co/MFK1z3S0Ag
VISION IA @vision_ia
22K Followers 202 Following Apprenez l'IA: https://t.co/fSoZsB1DzM J.Antoine Co-founder
Ansh Bhatnagar @iansh04_
7K Followers 38 Following 350K+ LinkedIn | Software Developer | AI & Tech Content Creator | Helping Brands to Grow | Open for Collaboration 👇 Contact: [email protected]
Modulo Labs @Modulo_Labs
104 Followers 61 Following Advancing scalable, privacy-preserving blockchain infrastructure with ZK for the real world. Bridging Web3, compliance & real-world utility | Home of @zk_monk
G.SKILL @gskillgaming
182K Followers 299 Following The world's leading brand of PC memory since '89. Game on. Game True. Game with Extreme Performance.
R Λ Z Ξ R @Razer
3.0M Followers 634 Following For Gamers. By Gamers. @RazerSupport will get you back in action. Join our official Discord: https://t.co/czT8GiB1wT
ROG Global @ASUS_ROG
1.1M Followers 341 Following The Official ASUS ROG Account. Join us on Discord https://t.co/5XXPrPYjEY For support, please go to https://t.co/xzem2yeEHg
Logitech G @LogitechG
1.6M Followers 2K Following Keep Playing 🦾 Official PC Peripherals of @callofduty Follow Logitech G: https://t.co/LjUHfl8MiS
MSI Gaming @msigaming
728K Followers 441 Following Tech Meets Aesthetic. #Truegaming Support: https://t.co/bgVQFyhh77
ZOTAC @ZOTAC
236K Followers 325 Following Welcome to the Official ZOTAC (Global) Twitter! #ZOTAC makers of #GraphicsCard, #miniPC & #gaming hardware. Follow us on Facebook: @ZOTAC & IG: @zotac_official
Intel Gaming @IntelGaming
1.3M Followers 428 Following Official Intel Gaming | Products, Events, Contests & Deals For product Qs or customer support, please visit Intel Support: https://t.co/xE448mOdMY
NVIDIA GeForce @NVIDIAGeForce
3.1M Followers 958 Following Introducing GeForce RTX 50 Series graphics cards & laptops, powered by NVIDIA Blackwell and AI.
ROG North America @ASUS_ROGNA
328K Followers 611 Following Official ASUS ROG (Republic of Gamers) Twitter of North America For ASUS NA: @ASUSUSA For RTX Graphics Cards: @ASUSGraphicsNA For product issues: @AsusHelpUS
Fireworks @FireworksAI_HQ
36K Followers 308 Following The frontier platform for training and inference on open-weights models at scale.
Together AI @togethercompute
65K Followers 276 Following Accelerate inference, model shaping, and pre-training on a research-optimized platform.
OpenAI Developers @OpenAIDevs
466K Followers 1 Following Official updates for developers building with Codex & the OpenAI Platform • Service status: https://t.co/kZwnwdYYEq
AMD Radeon @AMDRadeon
812K Followers 171 Following together we advance_gaming. Stay up to date with the latest tech and performance from @AMD Radeon™ graphics including the new Radeon™ RX 9000 Series graphics.
NVIDIA @nvidia
2.7M Followers 47 Following The official handle for NVIDIA. Blog: https://t.co/JAn5eKOTBT Support: https://t.co/6ln5FVnA2o All our social media: https://t.co/Uc56dL57Dh
Vast @vast
41K Followers 231 Following Building next-generation space stations and space infrastructure. Ensuring America’s continuous human presence in space.
Runpod @runpod
10K Followers 18 Following AI Developer Cloud. AI infrastructure developers trust. https://t.co/PWVZocfp6D
Hugging Models @HuggingModels
72K Followers 32 Following We're sharing/showcasing best of @huggingface models. Follow to stay in loop. Promoting Open-Source models.
Hugging Face @huggingface
829K Followers 227 Following The AI community building the future. https://t.co/TpiXQMQ9rZ
Mistral AI @MistralAI
225K Followers 2 Following Frontier AI in your hands. Get work done with @MistralVibe at https://t.co/JsGnCVMUFq.
ClaudeDevs @ClaudeDevs
786K Followers 2 Following Official updates for developers building with @ClaudeAI
Claude @claudeai
2.0M Followers 2 Following Claude is an AI assistant built by @anthropicai to be safe, accurate, and secure. Talk to Claude on https://t.co/ZhTwG8d1e5 or download the app.
OpenAI @OpenAI
5.5M Followers 4 Following OpenAI’s mission is to ensure that artificial general intelligence benefits all of humanity. We’re hiring: https://t.co/dJGr6LgzPA
Cartographe encarté ... @cartococo
22K Followers 4K Following Communiste humaniste. Liberté de circulation et d'installation pour toutes et tous. Prolétaires de tous les pays unissez-vous.
Civixplorer @Civixplorer
222K Followers 675 Following The eye of the world. Keep exploring → https://t.co/SdD5RzwxmbTrends for United States
You might like
