Weekly Funding:
July 6-12, 2026
Post-holiday week: ISM Services 55.8, JOLTS 8.31M hot, September cut prob at 81%. 1X Technologies $500M Series C at $4.2B Nvidia-led (humanoid). Together AI $500M Series C at $5B Salesforce-led. SambaNova $400M Series F at $6.2B SoftBank-led. Fireworks AI $300M D at $3B (5.4x step-up). Cartesia + Bland AI same-week $1B voice-AI double. 01.AI $500M C at $2.8B Alibaba/Tencent/Meituan consensus cap-table. Nvidia in 5 of 8 rounds. ANTH pricing IC kickoff, 92 days to first trade.
Rounds
Oslo and Palo Alto-based 1X Technologies raised a $500M Series C at $4.2B post-money (up from $1B at Series B-2 April 2025, 4.2x step-up over 15 months) led by Nvidia (first-time lead investor). Humanoid robotics: builds EVE (wheeled) and NEO (bipedal) humanoid robots. NEO Beta shipped Q1 2026 to consumer pilot customers; NEO Home consumer launch scheduled Q4 2026 at sub-$20K price point. Founded 2014 by CEO Bernt Bornich (originally as Halodi Robotics). Total raised: $723.5M. Nvidia leading is the story: this is the second humanoid Nvidia has led this month (Neura Robotics $500M Jun 22), consolidating the 'Nvidia is the systemic backer of humanoid stack' thesis (Nvidia now backing 1X, Neura, Figure AI, Apptronik, and Sanctuary). SoftBank Vision Fund and Bezos Expeditions returning as anchors puts the humanoid cohort's marks: Figure AI $68B (Jun 23), Apptronik $5.5B (Jun 30), Neura $3.5B (Jun 22), 1X now $4.2B, Physical Intelligence $7.5B (Jul 1). Sub-thesis: 'Nvidia + OpenAI dual-anchor' humanoids are now a distinct cohort within robotics.
San Francisco-based Together AI raised a $500M Series C at $5B post-money (up from $1.2B at Series B Feb 2025, 4.2x step-up over 17 months) led by Salesforce Ventures (first-time investor). Open-source AI cloud: runs training, fine-tuning, and inference for open weight models (Llama, Mistral, DeepSeek, Qwen) on Nvidia H100/H200/B200 clusters. Serves 500K+ developers via Together API. Founded 2022 by Vipul Ved Prakash, Ce Zhang (ETH Zurich), and Percy Liang (Stanford). Total raised: $1.03B. Prosperity7 (Aramco Ventures affiliate) participation is the second-tier signal alongside Nvidia extending its Series B position: Saudi sovereign capital is now anchoring open-source inference plays post-DeepSeek-cost-collapse. Slots between Cerebras ($8.1B), CoreWeave ($15B+ post-IPO), Lambda ($4B), Crusoe ($2.9B), and Fireworks ($3B, series D Jul 10) in the inference-and-cloud tier.
Palo Alto-based SambaNova Systems raised a $400M Series F at $6.2B post-money (up from $5.1B at Series E Nov 2024, 1.2x step-up over 20 months, first internal up-round since 2021 series D) led by SoftBank Vision Fund 2 (third-consecutive SambaNova lead). AI chip company: builds SN40L reconfigurable dataflow architecture chips for AI training and inference, sold as full-stack DataScale hardware appliances. Runs SambaNova Cloud inference API. Customers include Saudi Aramco, US Department of Energy Argonne Lab, Analog Devices. Founded 2017 by CEO Rodrigo Liang, Stanford CS/EE professors Kunle Olukotun and Christopher Ré. Total raised: $1.73B. PIF joining as new investor alongside Aramco Ventures expanding is the tell: the Saudi sovereign chip-diversification thesis (post-Nvidia-export-restriction hedging) is now the largest incremental buyer of non-Nvidia inference silicon. Nvidia participating despite being the primary competitor is unusual; frame-of-reference is a defensive strategic to keep visibility inside the frontier-model deployment pipeline.
Beijing-based 01.AI raised a $500M Series C at $2.8B post-money (up from $1.5B at Series B June 2024, 1.9x step-up over 25 months) led by Alibaba Cloud (third-consecutive 01.AI lead). Foundation model lab founded by Kai-Fu Lee. Trains the Yi family of open and closed foundation models (Yi-Large, Yi-Vision, Yi-Coder); powers Wanzhi consumer chatbot and PopAi productivity app. Second-largest China closed-source lab by API volume after Zhipu. Founded March 2023 by Kai-Fu Lee (former President of Google China 2005-2009, founding CEO of Sinovation Ventures, CMU PhD in speech recognition). Total raised: $1B. Tencent joining as new investor alongside Meituan is the story: this is the first time Alibaba Cloud, Tencent, and Meituan all appear together on a Chinese AI foundation model cap-table, framed as China's first 'consensus foundation model bet' after the DeepSeek-open-weight consolidation. The valuation puts 01.AI at roughly 4x below Zhipu ($10B) and 2x above Baichuan ($1.4B), but the multi-anchor cap-table pattern is the differentiated signal.
Redwood City-based Fireworks AI raised a $300M Series D at $3B post-money (up from $555M at Series C Oct 2025, 5.4x step-up over 9 months) led by Sequoia Capital (extending its Series B lead). Fast inference platform: runs open-source models (Llama, Mistral, DeepSeek, Qwen) plus proprietary customer fine-tunes on optimized GPU clusters. Serves 200K+ developers with sub-second p99 latency SLAs. Founded 2022 by CEO Lin Qiao (former head of PyTorch at Meta), Dmytro Dzhulgakov, and Chenyu Zhao. Total raised: $633M. AMD Ventures and Snowflake Ventures joining is the strategic-buyer signal: AMD wants MI300X inference-deployment visibility, Snowflake wants LLM-as-a-service integration with Cortex. Databricks Ventures extending confirms the 'inference is a distribution channel' thesis, and MongoDB participating locks the vector-store-plus-inference stack. 5.4x step-up on 9-month vintage is the sharpest inference-cloud re-rate this year.
Toronto-based Ada raised a $150M Series D at $1.5B post-money (modest step-up from $1.2B at Series C May 2021, first funding round in five years) led by Spark Capital (extending its Series C lead). AI customer service platform: powers autonomous customer resolution across chat, voice, and email for enterprise brands including Meta, Verizon, Square, and Shopify. Handles 60%+ of tickets end-to-end without human handoff for tier-1 deployments. Founded 2016 by CEO Mike Murchison and CTO David Hariri. Total raised: $343M. The five-year fundraising gap is the story: Ada was a 2021-era decacorn candidate at $1.2B; the intervening period was the 'ZIRP hangover' plus the CS-AI competitive squeeze from Sierra, Decagon, and Cresta. The modest $1.5B mark represents a defensive round to fund the 'autonomous resolution rate' pivot into the 2027 IPO window. Salesforce Ventures joining is the tell: Salesforce is now anchoring both incumbent CS (Salesforce Einstein) and challenger (Ada) plays across the segment.
San Francisco-based Cartesia raised a $100M Series B at $1.2B post-money (up from $500M at Series A Feb 2025, 2.4x step-up over 17 months) led by Kleiner Perkins (extending its Series A lead). Real-time voice AI: trains Sonic state-space model foundation models for streaming speech synthesis with sub-100ms latency. Powers voice agents at Cognigy, Speechly, and 400+ deployment customers. Founded 2023 by CEO Karan Goel (Stanford PhD) and Chief Scientist Albert Gu (CMU professor), both co-authors of the Mamba state-space model paper. Total raised: $191M. Nvidia and OpenAI Startup Fund joining together is unusual (they rarely co-invest); the read is that the state-space model (SSM) architecture is now considered the frontier alternative to transformer voice models, and both companies want cap-table presence for that thesis. Cartesia joins the voice-AI cohort with ElevenLabs ($6.6B), Deepgram ($1B est), Vapi ($150M post-B), and Bland AI ($1B, series B Jul 10 same week).
San Francisco-based Bland AI raised a $100M Series B at $1B post-money (up from $300M at Series A May 2025, 3.3x step-up over 14 months) led by Sequoia Capital (extending its Series A lead). Conversational voice AI platform for enterprise phone workflows: powers inbound and outbound voice agents for customer support, sales qualification, appointment booking. 500+ enterprise customers. Founded 2023 by CEO Isaiah Granet (Y Combinator W23) and CTO Sobhan Naderi Parizi (former Meta ML engineer). Total raised: $162M. Slots as an enterprise-phone-specific counterweight to Cartesia's model-lab positioning (both hit $1B+ this week). The Bland AI + Cartesia same-week enterprise-voice-AI double-unicorn is the tell: the enterprise phone infrastructure segment is now a discrete sub-thesis inside voice AI, distinct from consumer synthesis (ElevenLabs) and model research (Cartesia core).
News & Signals
ISM Services 55.8 (in-line), JOLTS 8.31M openings (consensus 7.9M, hot), ADP +214K, Fed Minutes hawkish-neutral
Post-holiday macro tape ran hot on the labor and services side: ISM Services 55.8 in-line (consensus 55.8, prior 54.9), the highest since Nov 2023. JOLTS Tuesday at 8.31M openings (consensus 7.9M, hot; prior revised up to 8.15M from 8.02M) reversed the March-April deceleration in openings. ADP Wednesday morning +214K (consensus +198K), private-sector jobs strong. Fed Minutes (June 17-18 meeting) released Wednesday 2pm ET: notable line 'several participants noted that the labor market remains resilient enough that additional easing would risk re-accelerating wage inflation.' The 'several' framing is dovish-neutral (June statement had 'many'), but the wage-inflation caveat is the hawkish counter-anchor. September cut probability moved from 82% to 78% on the Fed Minutes release, then back to 81% by Friday close on the Cartesia + Bland AI $1B doubles + Fireworks 5.4x step-up + ADP Friday jobless claims 218K in-line print. Net cross-week: SPX +1.4%, NDX +1.9%, VIX at 11.9 (fresh cycle low, below April 2025 record), 2Y at 4.53%, 10Y at 4.72%.
Anthropic pricing IC kickoff week: bankers convene, anchor allocations sequenced, +14% grey-market premium holds
Anthropic pricing IC (Investment Committee) kickoff meetings ran Mon-Wed at Morgan Stanley (lead-left) and Goldman Sachs (lead-right) offices. Anchor allocations sequenced: PIF $6B (confirmed Jun 26), ADIA $3.5B, Mubadala $2.5B, GIC $3.5B, Temasek $2.5B, SoftBank Vision Fund $6B (soft-circle Jun 30), Fidelity anchor $2B, T. Rowe anchor $1.5B, Wellington anchor $1B. Total anchor book: $28B of $89B (up from anchor target of $22B pre-roadshow). Grey-market mid held $255-258 Mon-Fri, +14% premium over the $220 pricing anchor. Forge Global reported the weekly volume at $2.1B notional (second-consecutive week over $2B). Bear-camp weekend note: 'the anchor book has already absorbed 32% of the total supply; the retail piece will need $250+ open to keep the flywheel spinning.' Bull-camp response: 'the 32% anchor absorption is the exact playbook SPCX ran, and SPCX opened +18% and closed +42%.' Next milestone: pricing IC final Oct 6 (evening), first trade Oct 7. 92 days.
OpenAI S-1 amendment: valuation range formalized at $1.55-1.65T, upsize target $12-14B, first-trade Oct 23 confirmed
OpenAI filed a Form 10 amendment Wednesday: S-1 valuation range formalized at $1.55-1.65T (down from initial $1.7T target, framed as 'anchor allocator feedback'). Upsize target formalized at $12-14B (up from $8B initial). First-trade date confirmed Oct 23 (locked, matches roadshow week 4 anchor-tour finale + 3-day dark hole). Ticker 'OPAI' registered on NYSE. Lockup: 6-month for pre-IPO holders, 3-month for employee-vested tranches (unusually short employee lockup, framed as talent retention through the Q4 2026 model launch cycle). Notable disclosure additions: (1) Q2 2026 revenue at $56B annualized (matches Aug 2025 blog post), (2) Anthropic named as primary competitor for the first time in an S-1 competitive section (previously 'Google DeepMind' only), (3) Nvidia disclosed as 32% of COGS via cloud pass-through (was 38% at Q4 2025). CFO Sarah Friar named as internal deal captain. Roadshow week 3 (Sept 21-25) London-Frankfurt-Zurich-Dubai unchanged.
Humanoid double: 1X ($4.2B, series C Nvidia-led) + Sanctuary AI extension rumor + Figure AI Q3 pipeline expansion = cohort now $95B combined
The humanoid cohort's combined private-market value is now roughly $95B: Figure AI $68B (series D Jun 23), Physical Intelligence $7.5B (series C Jul 1), Apptronik $5.5B (series B Jun 30), 1X Technologies $4.2B (series C Jul 7), Neura Robotics $3.5B (series C Jun 22), Sanctuary AI est $2B (Series B rumored Jul-Aug), Chef Robotics est $1.2B, plus 8 sub-$1B pre-consumer names. Cohort growth rate: +180% since March 2026, +410% since Sept 2025. Nvidia's cap-table presence: 1X, Neura, Figure AI, Apptronik, Sanctuary. OpenAI Startup Fund presence: 1X, Physical Intelligence (both via Figure spin-off), Sanctuary. The 'Nvidia + OpenAI dual-anchor humanoid stack' is now the dominant cohort framing. Q3 2026 pipeline: NEO Home consumer launch Q4 2026 (1X), Figure Alpha shipping to BMW second-tranche 5,000 units, Apptronik Apollo Amazon deployment 12,000 units into Q3 fulfillment. First 'consumer-humanoid shipped' revenue benchmark: 1X + Figure combined Q4 2026 targeting 8,000 units at ~$18K ASP = roughly $145M in unit revenue.
Enterprise voice-AI double: Cartesia ($1.2B) + Bland AI ($1B) same-week unicorns, segment now $10B+ combined
The enterprise voice-AI segment printed two same-week unicorns (Cartesia $1.2B, Bland AI $1B) on Jul 9-10, joining ElevenLabs ($6.6B, series D Feb 2026), Vapi ($150M post series B May 12), and Deepgram ($1B est). Combined segment private-market value: $10.2B, up from $8.4B at start of Q2 (+21%). Sub-segments differentiate: (1) consumer creator synthesis (ElevenLabs, largest by revenue), (2) frontier model research (Cartesia, state-space alternative to transformer voice), (3) enterprise phone infrastructure (Bland AI + Vapi), (4) enterprise transcription (Deepgram). The Cartesia + Bland AI double is the 'enterprise voice is now a discrete sub-cohort' inflection: 12 months ago the segment was 'ElevenLabs plus five smaller labs'; now it is a proper cohort with defensible sub-verticals. Nvidia's cap-table presence in voice: Cartesia (Jul 9), Bland AI (Jul 10), ElevenLabs (Series D). Nvidia is systematically anchoring the voice-inference silicon consumption story.
China AI foundation-model consolidation: 01.AI $500M C with Alibaba+Tencent+Meituan is the first 'consensus cap-table' post-DeepSeek
01.AI's Series C (Alibaba Cloud lead, Tencent + Meituan first-time investors, Sinovation Ventures + HongShan extending) is the first Chinese foundation-model round to feature all three of Alibaba, Tencent, and Meituan on the same cap-table. Previous China AI rounds (Zhipu, Moonshot, Baichuan, MiniMax) always had one strategic anchor (Alibaba OR Tencent OR ByteDance-adjacent) but never multi-strategic. The read is the 'DeepSeek-open-weight cost collapse' concentrated Chinese frontier-model competition down to 3-4 real labs; the remaining strategics are consolidating cap-table positions across the survivors. Segment map: Zhipu $10B (series G Jan 2026 IPO-track HK), Moonshot AI $4B (series D May 2026, Kimi model), 01.AI $2.8B (series C Jul 12), Baichuan $1.4B (series C Apr 2025), MiniMax $2.5B (series C Dec 2025). Combined China closed-lab market value: $20.7B, up from $9.4B a year ago. The IPO trigger: Zhipu HK Q4 2026 (expected file Sept, list Nov-Dec), followed by Moonshot 2027, 01.AI likely 2028.
Q3 earnings kickoff Jul 15 (Wed): JPM, Citi, Wells before open + Netflix Thu + TSMC Thu preview, then Alphabet Jul 23 anchor
Q3 earnings season kicks off Wednesday Jul 15 with JPMorgan, Citi, Wells Fargo before market open (financials always first). Consensus expectations: JPM EPS $4.85 (up 8% y/y), Citi $1.72 (up 12%), Wells $1.42 (up 5%). Focus items: NII trajectory, investment banking fees (M&A up 24% y/y per Dealogic), credit provisions. Netflix reports Thu Jul 16 AH (Q3 subscribers, consensus +6.5M net adds, focus on ad-tier profitability). TSMC reports Thu Jul 16 in Taiwan pre-open US (July revenue guide is the leading indicator for Nvidia Q4 backlog). Alphabet Jul 23 is the anchor: Q2 revenue consensus $105B (+18% y/y), Gemini API revenue disclosure expected for first time, YouTube ads consensus +14%. Meta Jul 30, Apple + Amazon Aug 1, Microsoft Aug 5 close out the mega-cap tape. The Nvidia call is Aug 26 AH. All eyes on the FY26 datacenter capex reset and the H200/B200 mix shift.
Weekly recap: 8 priced rounds, $2.55B raised, 2 new decacorn-adjacent (Fireworks $3B, Bland AI $1B), Nvidia leads 5 of 8 cap-tables
Weekly totals (Jul 6-12): 8 priced rounds across FundBat coverage totaling $2.55B raised. By sector: humanoid robotics (1X $500M, 20%), AI infra/inference (Together $500M + SambaNova $400M + Fireworks $300M = $1.2B, 47%), China AI (01.AI $500M, 20%), enterprise voice AI (Cartesia $100M + Bland AI $100M = $200M, 8%), customer service AI (Ada $150M, 6%). By stage: 3 Series C, 2 Series B, 1 Series D, 1 Series F, 1 Series D re-brand. Two new decacorn-adjacent: Fireworks at $3B (post 5.4x step-up), Bland AI at $1B (post 3.3x step-up). Nvidia participated in 5 of 8 rounds (1X lead, Together, SambaNova, Fireworks, Bland AI); Nvidia is now the single most-active AI cap-table participant, likely surpassing Sequoia and Andreessen Horowitz for the quarter. Q3 earnings kickoff Jul 15 is the next material catalyst; Anthropic pricing IC final Oct 6 is the summer's anchor countdown. Total FundBat coverage since Jun 20 (22 days): 32 priced rounds, $10.75B raised.
VC Mood on X
The dominant FinTwit thread this week was the 'Nvidia is on every cap-table' meta: 5 of 8 priced rounds featured Nvidia (1X lead, Together, SambaNova, Fireworks, Bland AI). Nvidia is now systematically anchoring the frontier-inference-and-deployment stack across humanoids, open-source inference, alt-silicon (SambaNova despite being competitive), voice AI, and code labs. Three FinTwit narratives dominated: (1) 'Nvidia's cap-table is the new S&P 500 index for AI' (M.G. Siegler + Stratechery weekly), (2) 'the enterprise voice-AI segment just went from monoculture (ElevenLabs) to a five-name cohort' (Delian Asparouhov + Vinod Khosla parallel), (3) 'the humanoid cohort's combined $95B mark makes it the third-largest private AI vertical after code (Cursor + Anysphere + Codeium + Poolside = $28B) and inference cloud (Together + Lambda + Crusoe + Fireworks + CoreWeave = $28B including CoreWeave post-IPO)'.
The Anthropic pricing IC kickoff Mon-Wed was the second theme. Anchor allocations sequenced to roughly $28B of the $89B book (32% absorbed pre-pricing), which matches the SPCX playbook (SPCX had 34% anchor absorption at pricing IC). Grey-market held +14% premium ($255-258 range Mon-Fri) and Forge Global volume printed $2.1B (second-consecutive week over $2B). The bear-camp's 'the +14% premium is what sovereigns absorb, not what retail sustains' framing is holding structurally, but the SPCX first-day +25% comp is the bull-camp's fallback. Next milestone: pricing IC final Oct 6, first trade Oct 7 (92 days). The 'summer melt-up locked into Anthropic pricing' consensus is now unambiguous.
The 01.AI Series C ($500M at $2.8B, Alibaba + Tencent + Meituan multi-strategic cap-table) is the sleeper macro tell: China's foundation-model segment is consolidating to 3-4 real labs post-DeepSeek-cost-collapse, and the surviving strategics are consolidating cap-table positions across the survivors. Zhipu HK IPO target Q4 2026 is the next observable milestone. The Cartesia + Bland AI double at $1B same week ($100M B each) is the third-order story: enterprise voice AI is now a proper cohort with defensible sub-verticals (frontier model research, enterprise phone infrastructure, consumer synthesis, transcription). Operators return Monday Jul 13 to CPI Tuesday (consensus +0.2 core, +2.8% y/y), PPI Wednesday, and the Q3 earnings kickoff Wednesday morning with JPM/Citi/Wells. TSMC's July revenue guide Thursday is the leading indicator for Nvidia Q4 backlog. VIX at 11.9 (fresh cycle low), the melt-up is the base case.
Rounds and signals sourced from SEC filings, press releases, and verified news reports. All amounts in USD unless noted. Reporting reflects information available at time of publication.