Together AI, a provider of AI-specialized cloud (neocloud) infrastructure, said on July 1, 2026 that it had raised $800 million in a Series C round, reaching a post-money valuation of $8.3 billion. Total funding raised now stands at $1.3 billion.
July 1, 2026 · Together AI
Together AI raises $800M, hits an $8.3B valuation
The open-source AI "Neocloud," co-founded by FlashAttention inventor Tri Dao, more than doubles its worth in a year as enterprises pivot to cheaper open-weight inference — with Nvidia and Saudi Aramco's venture arm backing the round.
$8.3B
Post-money valuation
$1.3B+
Total funding to date
500MW+
Compute capacity committed
Valuation more than 2.5× in one year
From $3.3B (Feb 2025) to $8.3B (Jul 2026).
Annual recurring revenue
$618M
→
~$1B
End 2025 → Feb 2026 (estimated)
Customer inference cost savings
6×–20×
Cheaper on open-weight models — Decagon cut costs 6× after migrating.
Round led by Prosperity7 Ventures (Saudi Aramco), with Nvidia backing the open-weight inference layer
A full-stack "AI Native Cloud" optimizing inference for open models — DeepSeek, Nemotron, MiniMax, Kimi, GLM — across H100 / H200 / B200 / GB200 GPUs.
Nvidia
Vista Equity Partners
General Catalyst
Emergence Capital
Schneider Electric
Salesforce Ventures
Pegatron · March Capital
STRENGTHS
Research-driven kernels (FlashAttention-4, Megakernel)
Hundreds of tokens/sec output speed
Comparable performance at lower cost
Used by Cursor, Eleven Labs, Suno
FRICTION POINTS
Unreturned authorization holds
Short billing intervals, slow refunds
"Fast but not revolutionary," some say
Dedicated endpoints advised at scale
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…