Claude Opus 4.7 and GPT-5.5 for UK SMEs: April 2026 deployment guide

April 2026 brought two frontier model releases inside a fortnight. Anthropic released Claude Opus 4.7 on 16 April 2026, with measurable gains on long-horizon agentic work, vision, and long-running coding tasks, plus task budgets in public beta as a first-class cost-control mechanism. OpenAI launched GPT-5.5 on 23 April 2026, the first fully retrained base since GPT-4.5, alongside a new Codex-only seat type on ChatGPT Business and Enterprise that took effect on 2 April 2026, when the standard ChatGPT Business seat price also reduced by 5 US dollars per month. The commercial implication for UK SMEs is not that one provider has won. It is that the deployment decision is now better expressed function-by-function and across both providers, rather than as a single vendor choice. This guide sets out what changed, where each model fits, what the total cost looks like in GBP at typical SME seat counts, and the UK compliance footnote that still has to be respected regardless of which model is chosen.

What is Claude Opus 4.7?

Claude Opus 4.7 is Anthropic's frontier model, generally available from 16 April 2026, designed for long-horizon agentic work, complex coding, and document reasoning. It is the successor to Claude Opus 4.6 and is available through claude.ai, the Claude API, AWS Bedrock (in US East N. Virginia, Asia Pacific Tokyo, Europe Ireland, and Europe Stockholm at launch), and Google Cloud Vertex AI with EU multi-region routing. API pricing is unchanged from Opus 4.6 at 5 US dollars per million input tokens and 25 US dollars per million output tokens, with prompt caching reducing repeated input to 0.50 US dollars per million and batch processing cutting standard rates by 50 per cent.

The change against Opus 4.6 is concentrated in three areas. First, long-horizon agentic performance: Anthropic's release notes describe the model as working coherently for hours on multi-step tasks rather than degrading after a few minutes, which matters for any agent loop a business actually depends on. Second, vision: Opus 4.7 accepts input images up to 2,576 pixels on the long edge, more than three times the previous limit, which improves comprehension of complex documents, charts, and multi-page PDFs. Third, the General Finance benchmark that Anthropic publishes alongside the release moved from 0.767 on Opus 4.6 to 0.813 on Opus 4.7. That is a meaningful gain on a finance-heavy task set, although as with any single benchmark a UK SME should still pilot it on its own actual workload before treating it as decisive.

Task budgets are the new cost-control feature most relevant to a finance-conscious buyer. They allow a developer or workflow owner to cap the token spend Claude is permitted to use across a long-running task, which prevents the most common failure mode in agentic systems: an unbounded loop that runs up a four-figure bill before anyone notices. Task budgets are in public beta at the time of writing.

One caveat is worth pricing in. Anthropic shipped a new tokenizer with Opus 4.7. Public reporting suggests the same source text can produce up to roughly 35 per cent more tokens depending on content type, so the headline price is unchanged but the effective cost per request can rise. Treat the published rates as a ceiling and run a sample workload before committing to volume.

What is GPT-5.5 and the new Codex seat?

GPT-5.5 is OpenAI's frontier model, released on 23 April 2026 and rolled into ChatGPT Plus, Pro, Business, Enterprise, Edu, and Go plans on launch. It is the first fully retrained base since GPT-4.5 rather than an incremental fine-tune, with omnimodal input that handles text, images, audio, and video in a single architecture. The standard API price is 5 US dollars per million input tokens and 30 US dollars per million output tokens, with cached input at 0.50 US dollars per million. A separate GPT-5.5 Pro variant is priced at 30 US dollars per million input and 180 US dollars per million output. Context is 400,000 tokens in the ChatGPT product surface and 1 million tokens in the API.

GPT-5.5's headline benchmark claim is Terminal-Bench 2.0 at 82.7 per cent. Terminal-Bench evaluates real command-line workflows, including planning, iteration, and tool coordination in a sandboxed terminal, and it is the closest published proxy for how an agent performs against a developer's daily working environment. Anthropic reports Claude Opus 4.7 at 69.4 per cent on the same benchmark, so GPT-5.5 currently leads that specific test by more than thirteen points. The practical implication is narrower than the gap suggests, because Terminal-Bench measures a particular slice of agentic coding rather than enterprise document or finance work, but it is a real signal for terminal-heavy and codebase-heavy workflows.

The structural change for UK business buyers is the new ChatGPT seat model that took effect on 2 April 2026. ChatGPT Business and Enterprise workspaces now support two seat types: a standard ChatGPT seat with a fixed cost per user per month for typical knowledge-work users, and a Codex seat that is usage-based with no fixed monthly cost, suited to engineers who consume Codex agents and the command-line developer agent rather than the chat interface. The standard ChatGPT Business seat price also dropped by 5 US dollars per month on the same date, settling at 20 US dollars per seat per month on annual billing and 25 US dollars per seat per month on monthly billing.

The Codex seat is the change a CFO will care about most. Engineers who previously consumed Codex tokens against a shared seat now sit on a usage-based contract billed against a Codex Rate Card aligned to token consumption. For a development team running heavy agent loops, this is a more honest pricing model. For a non-engineering team buying Business seats for general AI use, nothing changes except the price reduction.

Function-by-function fit

Single-vendor framing breaks at first contact with how UK SMEs actually use AI. Different functions have different ratios of long-horizon reasoning, vision, code, omnimodal input, and cost sensitivity. The table below maps the six functions where The AI Consultancy's clients deploy AI most often, against which model is the better fit at the time of writing and why.

Function	Better fit	Why	UK SME deployment note
Finance month-end	Claude Opus 4.7	Higher General Finance benchmark (0.813 against 0.767 on Opus 4.6); stronger long-document reasoning across reconciliations and multi-period accounts.	Most useful via the API or claude.ai with explicit non-training settings on a Team or Enterprise tier. Pair with task budgets to cap a long reconciliation run.
Contract and policy review	Either, with a tilt to Claude	Both handle long contracts well. Claude's vision improvement helps with scanned PDFs and printed contracts. GPT-5.5 wins on omnimodal cases that include audio (recorded calls, voice notes).	Most UK firms already pilot legal review on a single platform; switching providers per matter creates more friction than it removes.
Sales follow-up and email drafting	GPT-5.5	Omnimodal input, broader Connectors ecosystem, Custom GPT support across a sales team without engineering work.	Standard ChatGPT Business seat at 20 US dollars per user per month annual is the natural fit.
Internal data Q&A (RAG)	Either	Both serve the model layer of a RAG architecture well. Choice is usually decided by what the rest of the stack already integrates with rather than the model itself.	Pilot with the provider whose enterprise admin and DPA are already in place. Switching is cheap if the retrieval layer is owned.
Operations playbooks and runbooks	Claude Opus 4.7	Long-horizon coherence over multi-hour playbook execution; task budgets give a clean cap on agentic spend.	Useful for finance ops, compliance ops, and back-office automation. Always require human approval at irreversible steps.
Knowledge work (drafting, summarisation, research)	Either, with a tilt to GPT-5.5 for breadth and Claude for depth	GPT-5.5's omnimodal input and broader plug-in ecosystem suits casual users. Claude's longer outputs and document quality suit specialist analysts.	Most UK SMEs pick one for the firm-wide standard and let specialists request the other. Price difference at this scale is noise.

The honest reading of the table is that Claude is the stronger choice for finance, ops, and long-document work, GPT-5.5 is the stronger choice for sales, omnimodal cases, and broad knowledge-work surface area, and the rest are draws decided by existing stack integrations. For a longer-form companion to this analysis, see our Claude vs ChatGPT Enterprise for UK SMEs comparison, which sets out the prior-generation Opus 4.6 and GPT-5 comparison and the underlying total-cost-of-ownership method.

Cost framing in GBP at 10, 50, and 200 seats

All figures below are exclusive of VAT and stated as of the date of writing. Where a number cannot be verified on a vendor's live pricing page it is marked indicative and the assumption is shown. UK VAT at 20 per cent applies to all subscriptions for UK business customers and is recoverable in the normal way for VAT-registered firms. GBP figures are at an indicative rate of approximately 0.79 GBP to 1 US dollar; finance teams should refresh the rate against the actual procurement-day rate before signing.

Claude Team is 20 US dollars per seat per month on annual billing, equivalent to roughly 15.80 GBP per seat per month, with premium seats at 100 US dollars (about 79 GBP). Claude Enterprise is published at 20 US dollars per seat plus usage at API rates, so a faithful total cost requires a usage projection in addition to the seat fee. ChatGPT Plus is 20 US dollars per month per individual subscriber, equivalent to roughly 15.80 GBP net of VAT. ChatGPT Business at 20 US dollars per seat per month annual is also approximately 15.80 GBP, with the new Codex seat type billed on usage rather than a fixed seat fee. ChatGPT Enterprise is custom-priced and typically in the 45 to 75 US dollar per seat per month range, with a 150-seat minimum, so it is genuinely not available below that floor.

At 10 seats, the seat-fee comparison is essentially flat. Claude Team or ChatGPT Business at 20 US dollars per seat per month gives an annual seat cost of roughly 1,900 GBP either way. Choose on compliance, integrations, and the function-fit table above, not on the seat price.

At 50 seats, the same arithmetic gives an annual seat cost of roughly 9,500 GBP on either platform. Claude Enterprise is the additional option at this scale, with the 20 US dollar seat fee giving a similar floor but with API usage layered on top. A team that uses Claude Projects heavily can add a few thousand pounds a year of API consumption; a team that runs heavy agent loops without task budgets has historically been able to add ten thousand or more. This is exactly the bill-surprise risk that task budgets are intended to remove. ChatGPT Enterprise is still not usually available below 150 users.

At 200 seats, ChatGPT Enterprise comes onto the table at scale, with negotiated rates typically in the 45 to 60 US dollar range per seat per month for buyers committing to multi-year terms, equivalent to roughly 85,000 GBP to 115,000 GBP per year before any premium tooling, training, or admin overhead. Claude Enterprise at the published 20 US dollar seat fee is roughly 38,000 GBP per year before usage, materially cheaper on the seat line, but the usage layer still has to be modelled honestly. For most UK SMEs at 200 seats, the right move is to model both options on the actual workload mix from a 30-day pilot before committing.

Decision checklist

Five questions resolve most procurement decisions in 2026. The headlines are the same as twelve months ago; the answers have moved.

Pick Claude as the firm-wide standard when the dominant use case is long-document reasoning, finance work, ops playbooks, or any agent loop with a multi-hour execution window. Pick ChatGPT as the firm-wide standard when the dominant use case is sales, marketing, omnimodal customer interaction (audio, video), or broad knowledge work across many casual users. Run both when the firm has functionally distinct teams (for example, finance ops on Claude and sales on ChatGPT) and the admin overhead is not a binding constraint. Use the API directly via AWS Bedrock (Europe Ireland or Stockholm for Opus 4.7 at launch) or Google Cloud Vertex AI EU multi-region when the workload is automation-heavy rather than chat-heavy, when seats would be a wasteful price model for the actual usage, or when a stricter EU residency boundary is required than the consumer surfaces provide.

The most common procurement mistake in 2026 is treating the seat price as the comparison axis. At 10 to 50 seats, the seat price is within procurement noise. At 200 seats it dominates the seat line but is still not the binding constraint, because usage costs and admin overhead can swing the total by more than the seat differential. Make the comparison on function fit, compliance fit, and total cost across seats and usage, not on which seat is a few pounds cheaper.

UK compliance footnote

The compliance baseline does not move because the models did. Three rules still apply, regardless of provider.

Under UK GDPR, any processing of personal data by an AI vendor on the firm's behalf requires a Data Processing Addendum compliant with Article 28, executed before the data is sent. Both Anthropic and OpenAI offer DPAs on business and enterprise tiers. The consumer tiers do not, which is why the consumer-tier ban is non-negotiable for any business workflow involving personal data, even for individual experimentation by employees.

Data residency is a partial story. ChatGPT Enterprise added native UK data residency at rest on 24 October 2025, which is the cleanest residency story for UK personal data on the OpenAI side. Claude Opus 4.7 on AWS Bedrock is available at launch in EU Ireland and EU Stockholm, and on Google Cloud Vertex AI through an EU multi-region endpoint; both are EU residency rather than strict UK residency. The direct claude.ai surface still processes primarily in the United States, so for regulated-sector UK SMEs the operational answer is either ChatGPT Enterprise on UK residency or Claude through Bedrock or Vertex AI in EU regions. For sectoral context, see our financial services industry page and our professional services industry page.

Two adjacent rules round it out. ICO guidance on AI and data protection still applies, including the requirement to complete a Data Protection Impact Assessment for higher-risk processing. The EU AI Act's high-risk obligations bind UK firms with EU data subjects from 2 August 2026, regardless of which model they use behind the scenes.