How to Trade Cerebras Systems (CBRS) Stock 24/7 with Free Trading Bots on Hyperliquid

Disclosure: Fomoed may earn a small commission if you open an account through the exchange links in this article.

Cerebras Systems (CBRS) is the most volatile new public ticker in the entire AI infrastructure complex — and one of the more polarizing. The May 2026 IPO was the largest semiconductor IPO of all time: $6.4 billion raised, priced at $185, opened at $350, closed the first day at $311. Six weeks later the stock had crashed roughly 47% from the open on margin-compression and customer-concentration concerns. Q1 2026 revenue (the first report as a public company) came in at $193.4 million — up 92% year-over-year — but management warned core gross margin would shrink to 36–38% in Q2 from 46.5% in Q1. The OpenAI deal (multi-year, 750MW, valued at more than $20 billion) and the Amazon AWS partnership give the bull case an enormous structural floor. The wafer-scale technology genuinely is the fastest AI inference on the planet. But the path from "genuinely fastest" to "durably profitable" is the whole bet.

The catch, as always with newly-public U.S. equities, is the cage. CBRS trades on the Nasdaq, which means six and a half hours a day, Monday through Friday, minus holidays. Every catalyst that actually moves Cerebras — an OpenAI deployment milestone, an AWS instance launch, a competitor benchmark (Groq, SambaNova), a sales-cycle update from Tan or anyone on the executive team — lands either after the bell, before the open, or at random points in the developer-conference calendar. By the time the U.S. market reopens, the move has happened and retail is already trading the wake.

Hyperliquid now lists a xyz:CBRS perpetual contract that trades 24/7 with isolated leverage. Long or short, with the same USDC margin you use for BTC, NVDA, and SPX. No broker, no PDT rule, no after-hours liquidity cliff. Combined with Fomoed's free DCA, grid, and custom strategy bots, retail traders finally have an automated, around-the-clock, no-KYC path into the most-watched new AI IPO of 2026.

Why CBRS Trading Hours Are Broken for Retail

Like every U.S.-listed name, Cerebras is locked inside the same 6.5-hour Nasdaq window we covered in our companion piece on trading NVDA 24/7 on Hyperliquid. Regular hours run 9:30am–4pm Eastern. Pre-market and after-hours sessions exist on most retail brokers (4am–9:30am and 4pm–8pm ET), but for newly-listed names like CBRS the liquidity outside cash hours is dramatically worse than for mega-caps. Spreads can widen to 100+ basis points in extreme cases. Many brokers restrict stop-losses or block all but limit orders. Outside of those 16 hours of partial liquidity, CBRS simply does not trade for retail.

Cerebras is particularly punishing for off-hours catalysts because, as a brand-new public company, almost every catalyst is unscheduled and happens whenever the announcement is ready:

• Customer-deployment milestones. The OpenAI 750MW deal is multi-year. Every operational milestone (data center go-live, capacity additions, model-deployment-on-Cerebras headlines) is a single-day catalyst that lands at whatever time the customer publishes.
• AWS instance launches. Cerebras and AWS announced a multi-year partnership to bring Cerebras inference to AWS. Every new instance availability, every region rollout, every public benchmark moves the tape.
• Competitive benchmarks. Groq, SambaNova, Tenstorrent, and increasingly NVIDIA's own software optimizations (TensorRT-LLM) all publish benchmarks. Each one is a CBRS-tape event because the entire bull case hinges on Cerebras being demonstrably the fastest inference platform.
• Margin guidance updates. The Q1 guidance to 36–38% margins in Q2 (down from 46.5%) was the single largest CBRS event since IPO. The stock dropped 28% in a session. Any updates on the margin trajectory will move the stock similarly.
• Lock-up expirations. Insider lock-up expirations are scheduled supply events that explicitly compress prices into the calendar date. Knowing the dates is half the battle.

The structural result: CBRS has been the most volatile newly-public AI ticker of 2026. Single-day moves of 15–30% are routine. Retail traders relying on brokers either eat the gap or skip the trade. Hyperliquid's 24/7 perp closes the gap problem.

What Cerebras Actually Builds (and Why OpenAI Pays)

To trade CBRS with any conviction you have to understand what they actually sell. Cerebras built and patented the Wafer-Scale Engine (WSE) — a single AI accelerator chip that takes up an entire 12-inch silicon wafer. The WSE-3 (current generation) is roughly 56× larger than the largest NVIDIA GPU and integrates approximately 4 trillion transistors with 900,000 AI-optimized cores on a single die. The architectural advantage is bandwidth: by keeping the entire model on a single wafer, Cerebras eliminates the inter-chip communication latency that dominates training and (especially) inference workloads on conventional GPU clusters.

1. Wafer-Scale Engine + Cerebras Inference Service

The headline benchmark — Cerebras consistently posts the fastest published latency and tokens-per-second numbers for inference workloads, often by an order of magnitude over comparable NVIDIA H100 clusters. The Cerebras Inference Service is the productized version: a managed API that customers call to run inference workloads at high speed. It is the franchise.

2. The OpenAI Deal

The most consequential single fact about Cerebras is the OpenAI partnership: multi-year, 750MW of dedicated capacity, valued at more than $20 billion over the contract life. The strategic logic: OpenAI needs inference cost-per-token to come down faster than NVIDIA's product roadmap delivers, and Cerebras's wafer-scale latency advantage is a differentiated tool in that effort. Whether OpenAI is using Cerebras for ChatGPT inference, for research workloads, or as a hedge against NVIDIA pricing is the open question — but the contract is signed.

3. The AWS Partnership

The AWS deal makes Cerebras inference available as managed instances on AWS — which is a category-defining moment for a non-NVIDIA AI silicon vendor. Until now, AWS has been one of NVIDIA's largest customers; offering Cerebras instances is an explicit hedge by Amazon against single-vendor risk. Each new region rollout and each new instance type is a Cerebras catalyst.

The Margin Problem

The bear case is the margin. Q1 2026 core gross margin was 46.5%, which is below where NVIDIA has run AI silicon margins for years. Q2 guidance for 36–38% was the surprise that crashed the stock 28% in a single session. The pattern is consistent with explosive revenue growth in low-margin large-customer deployments — the OpenAI deal in particular is likely structured at low margin in exchange for volume — but the buyside is rightly cautious about how durable the margin recovery is.

What this product mix means for traders: CBRS is the highest-conviction-with-highest-uncertainty AI hardware story on the board. The technology is genuinely differentiated. The flagship customer (OpenAI) is the most-watched AI buyer in the world. The financial trajectory is unproven. The float is small. Volatility is dramatically higher than any mega-cap — perfect for grid and DCA strategies that profit from chop, but demanding on position sizing.

The 2026 Setup: Why Now Is Different

Three concrete catalysts will likely set the CBRS range through year-end:

1. Q2 print (likely late August). The number that matters is core gross margin. If Cerebras prints in the 36–38% range as guided and management raises guidance for Q3, the multiple re-rates. If the print comes in below 36% or guidance is taken lower, the post-print drawdown could be another 20–30%.
2. OpenAI deployment milestones. Any specific update on the 750MW capacity ramp, deployment timing, or workload mix is a tradeable headline. The buyside has effectively zero visibility today.
3. Lock-up expiration calendar. The IPO lock-up tends to expire 180 days after the listing. The selling pressure into and out of those dates is a known supply event and trades like one.

Trading all three catalysts means being able to react to news that lands at 4:30pm ET, at random points in the OpenAI press cycle, and at the exact close of a calendar date. That's exactly what 24/7 perp access buys.

Setting Up a Cerebras Trading Bot on Hyperliquid

Three free strategy templates inside Fomoed are well-suited to CBRS's price action — but the position sizing rules apply with extra weight here because CBRS volatility is in a different league.

Strategy 1 — DCA the Drawdowns

CBRS is the textbook case for a DCA bot precisely because the volatility is so high and the structural support (OpenAI deal, AWS partnership, technology differentiation) gives drawdowns a probabilistic floor. A DCA bot ladders into the position at predefined drawdown levels.

• Base order: $25–$100 — sized very small relative to your normal positions, because CBRS's volatility means full-ladder drawdowns can exceed 50%.
• Safety orders: 6–9 layers — very wide spacing (2–2.5× step factor) because CBRS drawdowns are wider than any other AI name.
• Take profit: 5–8% from the average entry — CBRS rarely round-trips a full drawdown without an interim 6–10% bounce.
• Trigger filter: RSI < 25 on the 1-hour chart (yes, lower than mega-caps because CBRS routinely prints sub-30 RSI), with a 200-EMA trend filter on the 4-hour to skip drawdowns that are part of a narrative collapse.

The full DCA walkthrough generalises to CBRS — just remember to be more conservative on base order than you would for mega-cap names.

Strategy 2 — Grid the Post-Crash Range

After the 47% drawdown from the IPO highs, CBRS has been trading in a 30–45% range. That's textbook grid territory, but the rungs need to be wide.

• Range: upper bound at the post-IPO consolidation high; lower bound at the post-Q1-print low.
• Levels: 12–16 grid lines geometric — CBRS's high absolute volatility makes geometric spacing essential.
• Per-level size: 1/N of total grid capital where N is the level count.
• Stop-loss / unwind: hard floor 15% below range bottom; if CBRS breaks that, the grid unwinds and you reset after the next catalyst.

Our free grid bot handles arithmetic, geometric, and Fibonacci spacing.

Strategy 3 — Custom RSI/EMA Reversion for News Extremes

CBRS over-extends on news in both directions because the float is small and retail / hedge fund positioning swings rapidly. A simple mean-reversion custom strategy bot works, but the parameters need to be tighter than for mega-caps:

• Entry long: RSI(14) crosses up through 25 on the 4-hour chart while the 50-EMA > 200-EMA.
• Entry short: RSI(14) crosses down through 75 while 50-EMA < 200-EMA.
• Scale-out TP1: 50% at touch of the 12-EMA.
• Scale-out TP2: remaining 50% at touch of the 45-EMA.
• Stop loss: 5–7% beyond the entry candle's extreme; move to break-even on TP1 fill.

Backtest the template inside Fomoed's free backtest sandbox before going live — and use a 6-month rather than 12-month window since CBRS only has a couple months of public-market trading history.

The AI-Inference Pair Trade: Long CBRS / Short NVDA (Beta-Adjusted)

The most interesting institutional pair on CBRS is long CBRS / short NVDA on a heavily beta-adjusted basis. The thesis is the bear case for NVIDIA: that as inference becomes a larger share of total AI compute spend (the trajectory most buyside research now agrees on), the part of the market where Cerebras is competitive expands. Cerebras's wafer-scale latency advantage is hardest to replicate exactly in the inference workload.

On Hyperliquid the implementation is two perp orders in the same isolated-margin account, with the size ratio heavily tilted because of CBRS's beta:

• Long xyz:CBRS, $3,000 notional, 2× leverage.
• Short xyz:NVDA, $10,000 notional, 2× leverage.

The 3:10 ratio reflects CBRS's ~3× beta to NVDA. Net delta to AI compute demand is roughly zero. Funding cost is the difference between the two funding rates. P&L is the relative move between the two legs.

A pure relative-value pair is long CBRS / short SMCI — both are high-beta AI compute names but Cerebras has the OpenAI customer concentration story and Supermicro has the NVIDIA OEM exposure. The correlation is loose enough that the pair carries real edge in either direction.

You can automate either pair with two bots running opposite directions on the same Fomoed account. The 24/7 access is the killer feature: when an OpenAI deployment headline lands at midnight, both legs move on Hyperliquid simultaneously.

Fee Math — Broker vs Hyperliquid Perp

Take a $10,000 CBRS round-trip with three intra-week trades:

Cash broker (Robinhood, Schwab, Fidelity):

• Commission: $0 per trade on most U.S. retail brokers.
• Spread + PFOF impact: CBRS spreads are wide — newly-public, small float. ~5–10 bps per side in regular hours; 20–50 bps in pre-market/after-hours. Three round-trips averages ~10 bps per side = ~60 bps = ~$60.
• Borrow cost on the short leg: CBRS borrow has been intermittently hard-to-borrow, with rates spiking to 15–30% annualised post-IPO. A one-week short = ~$40.
• After-hours liquidity restrictions: CBRS's biggest moves are mostly uncatchable on cash brokers. Cost = opportunity, and it's a big one.
• PDT rule: if your account is < $25,000 and you hit four day-trades in five rolling days, account frozen 90 days.

Hyperliquid perp (xyz:CBRS):

• Taker fee: 3.5 bps per side = 21 bps for three round-trips = ~$21.
• Maker fee: 1 bp per side or rebated, if you limit-order some legs = ~$10–15.
• Funding rate: CBRS funding has been the highest of any new AI ticker post-listing — peaks above 3 bps per 8 hours on one-sided positioning. One week of one-sided positioning = ~60 bps = ~$60. Symmetric: short the heavy side, receive instead of pay (this has been a real edge on CBRS).
• Borrow: not a thing on perps.
• PDT rule: not a thing on perps.
• After-hours: not a thing — perp trades 24/7.

For CBRS, the Hyperliquid math is dramatically better than the cash broker math on borrow alone — never mind the after-hours optionality. The funding rate is the equivalent risk on the perp side, and it can spike fast. Use the perp tactically, not as a permanent long-hold replacement.

The risk to flag: funding rate spikes. CBRS funding has periodically blown through 0.08% per 8 hours on one-sided positioning. Annualised that's a 90%+ cost-of-carry. The fix is to use the perp for event trades and to take the under-crowded side when funding gets extreme — the funding-rate-receive trade has been profitable in itself on CBRS.

Risk — What Actually Moves Cerebras Stock

If you're deploying capital on CBRS, know what catalysts to expect:

• Quarterly earnings + gross margin. The margin trajectory is the single most-watched metric. Each print is a major catalyst until management establishes a durable margin floor.
• OpenAI deployment milestones. Any data center go-live, capacity addition, or specific workload-on-Cerebras headline is a tradeable event.
• AWS instance launches. New region availability, new instance types, public benchmarks vs NVIDIA on AWS — each is a single-day catalyst.
• Competitive benchmarks. Groq, SambaNova, Tenstorrent, NVIDIA TensorRT-LLM all publish benchmarks. Each one moves the CBRS thesis on the speed lead.
• Lock-up expiration dates. Insider supply events are mechanical and tradeable. Know the calendar.
• Customer concentration disclosures. The Q1 print made customer concentration risk an explicit market concern. Any disclosure about non-OpenAI customer wins or churn is high-impact.

The pattern across all six: catalysts cluster outside U.S. cash hours and produce gaps that are uncatchable on cash brokers. A 24/7 perp turns those gaps into tradeable events — and on CBRS specifically, the gaps are the largest in the AI complex.

Getting Started in 5 Steps

1. Open a Hyperliquid account. Use this referral link to get a fee discount. The whole signup is < 2 minutes — connect a wallet, fund USDC, done. No KYC, no broker forms.
2. Connect Hyperliquid to Fomoed. In the dashboard, add Hyperliquid as an exchange. We use a builder code so the bot can sign trades on your behalf without ever holding your funds.
3. Backtest your strategy. Use the free backtest sandbox with pair xyz:CBRS and the longest available window (CBRS has only been public for ~2 months, so backtest data is short — use it as directional signal, not statistical truth).
4. Deploy a very small live bot. Start with $25–$100 position size. Use isolated leverage 1× while you verify the strategy. CBRS's volatility means you should size dramatically smaller than any other position in your book.
5. Add notifications + monitoring. Telegram alerts on every open/close and a daily P&L summary. Watch it for at least 3 weeks before scaling up.

Tax and Regulatory Notes

Hyperliquid is a non-custodial perpetual DEX. It does not collect KYC, does not issue 1099s, and does not custody funds — your USDC stays in your own wallet, the bot signs orders via a builder code. The user agreement excludes a short list of sanctioned countries; for everyone else, the venue is open. The flip side: you are responsible for tracking and reporting your own trades. In the U.S., perp P&L is typically treated as ordinary or §1256 income — talk to your accountant. Outside the U.S., crypto perp treatment varies wildly. 24/7 access is the upside, self-reporting is the cost of admission.

Conclusion — Cerebras 24/7 Is a High-Risk, High-Convexity Trade

CBRS is unlike any other ticker in your AI book. The technology is genuinely differentiated — wafer-scale inference is faster than NVIDIA's stack for the workloads it serves. The flagship customer is OpenAI. The AWS partnership is real. None of that is in question. What is in question is the margin trajectory, the durability of the OpenAI revenue concentration, and the path to GAAP profitability. The 47% drawdown from the IPO highs is the market priced in those uncertainties.

What 24/7 access changes is purely tactical. You can react to the Q2 print at 4:30pm ET. You can short into a competitive benchmark release at 11pm ET. You can long ahead of a lock-up expiration end-of-day. You can run the CBRS/NVDA pair as a single perp basket without two brokerage accounts. None of those alone is revolutionary, but compounded over a year on a name with CBRS's volatility, they are the difference between catching the move and reading about it on X the next morning.

Trade it with respect — CBRS is the most volatile AI ticker in mid-cap, the perp funding rate can be brutal, and the lock-up calendar adds known supply pressure. Use small size, isolated margin, and let the bots do the unglamorous work of laddering, gridding, and mean-reverting around the catalysts. That's the only sustainable way to trade a newly-public name with this kind of vol profile.

Related Resources

• How to Trade NVDA Stock 24/7 on Hyperliquid
• How to Trade AMD Stock 24/7 on Hyperliquid
• How to Trade Marvell (MRVL) Stock 24/7 on Hyperliquid
• Hyperliquid Trading Bot Setup Guide
• Trade SPX 24/7 with Free Bots on Hyperliquid
• AI Crypto Trading Bots Explained
• Free Grid Trading Bot on Hyperliquid
• How to Set Up a Free DCA Bot
• Free Grid Trading Bot
• Custom Strategy Bot
• Free DCA Trading Bot
• Start Trading Free