DeepSeek V4 approaches: 1M tokens of context and double pricing
DeepSeek is preparing for the public release of its V4, with a 1M token window and a double pricing grid (peak/off-peak). The shadow extends beyond a simple update.
yesterday 10 8
Oracle's credit downgrade deepens the AI-capex debt question
Tech in Asia flags rising Oracle credit risk as AI spending balloons. The BBB- warning is turning into a re-pricing.
yesterday 12 8
Kimi K3 pricing: China's frontier goes premium, ends the race-to-the-bottom
Moonshot has priced Kimi K3 at 3-15× the tokens of DeepSeek - the first time a Chinese frontier lab has publicly asked for a premium against its domestic peers, and a break with the price-war narrative.
Jul 18, 2026 at 11:11 20 9
Kioxia -16%, SoftBank in retreat: the Asian AI complex prices a deleveraging
Deleveraging AI in Asia: Kioxia drops 16%, SoftBank slides. The repricing of AI risk begins to show in the prices.
Jul 17, 2026 at 09:21 20 10
TSMC raises 2026 revenue forecast again: AI demand continues to set the baseline
TSMC raises its 2026 forecasts again, driven by AI demand. The compute crunch is no longer a hypothesis: it's the new pricing baseline.
Jul 17, 2026 at 09:20 12 9
Anthropic prices Claude in rupees for India: pricing localisation as the first act of sovereignty
Anthropic added rupee pricing for Claude in India. Small operational change, real strategic meaning - the fight for the Indian market is now about payment rails and taxation, not model quality.
Jul 16, 2026 at 08:48 30 9
Microsoft repositionne Copilot face à OpenAI et Anthropic : la phase de l'ami-ennemi se termine
Microsoft is now openly pitching Copilot against OpenAI and Anthropic, per Tech in Asia (16 July 2026). After the Build 2026 orchestration layer, Redmond ends the wrapper era and prices its own model story as the differentiator.
Jul 16, 2026 at 08:46 28 9
Meta ships Muse Spark 1.1 with a cheap Model API - the third-runner squeeze play
Meta released Muse Spark 1.1 (multimodal reasoning) and simultaneously priced its Meta Model API aggressively low. Neither will beat GPT-5.6 or Claude on benchmarks. That's not the point. The point is undercutting the meter.
Jul 11, 2026 at 11:41 29 8