Meta tags:
description=Claude Opus 4.7 review: Anthropic s new flagship leads SWE-bench Pro and MCP-Atlas. See what s new vs. Opus 4.6, GPT-5.4, Gemini 3.1 Pro, and Mythos Preview.;

Headings (most frequently used words):

claude, opus, transform, what, benchmarks, and, webkit, more, is, flex, none, features, tests, new, model, vs, gpt, for, on, agentic, in, the, shrink, ms, 18px, rotate, 5turn, translate, 21, 10, transition, 3s, cubic, bezier, 85, 15, hands, which, best, access, use, css, people, with, how, to, about, are, instruction, following, multimodal, reasoning, xhigh, effort, level, negative, height, padding, top, 6px, moz, width, agents, tools, tasks, gemini, frontier, should, you, sonnet, text, decoration, ai, anthropic, available, grouptraining, testing, pricing, else, launched, alongside, saying, conclusion, faqs, grow, your, data, skills, datacamp, mobile, support, memory, coding, visual, tool, computer, financial, analysis, ultrareview, code, auto, mode, max, users, task, budgets, api, mythos, preview, 18x2vi3, project, glasswing, 167dpqb, much, does, cost, performance, improvements, 1531qan, color, inherit, software, engineering, implementing, solutions, business, introduction, models,

Text of the page (most frequently used words):
the (160), and (107), opus (100), #claude (54), for (52), that (41), data (36), model (36), with (33), #anthropic (33), you (29), which (26), more (24), this (23), gpt (22), use (21), courses (21), what (20), datacamp (19), are (19), benchmarks (18), new (18), beta (17), not (16), about (16), models (16), high (16), but (16), coding (15), your (15), tests (15), pro (15), from (15), one (15), agentic (14), reasoning (14), mythos (14), our (13), tasks (13), than (13), can (13), all (12), across (12), business (11), gemini (11), has (11), also (11), every (11), pricing (10), blog (10), learn (10), see (10), how (10), now (10), better (10), following (10), preview (10), release (10), first (10), bench (10), scores (10), line (10), code (9), between (9), over (9), when (9), any (9), have (9), word (9), client (9), paragraph (9), science (8), sonnet (8), context (8), long (8), tool (8), frontier (8), its (8), best (8), benchmark (8), tools (8), work (8), level (8), test (8), swe (8), where (8), read (8), count (8), support (7), min (7), features (7), access (7), latest (7), computer (7), real (7), right (7), improved (7), resolution (7), instruction (7), memory (7), per (7), tokens (7), internal (7), users (7), visual (7), closing (7), briefing (7), must (7), category (7), partner (6), engineering (6), google (6), token (6), performance (6), compare (6), should (6), hands (6), agents (6), results (6), course (6), before (6), vision (6), does (6), available (6), above (6), their (6), was (6), score (6), chart (6), each (6), words (6), get (5), tom (5), workflows (5), api (5), software (5), intelligence (5), max (5), xhigh (5), stronger (5), same (5), output (5), back (5), approach (5), several (5), there (5), only (5), agent (5), higher (5), instructions (5), different (5), against (5), published (5), both (5), here (5), improvement (5), still (5), confidence (5), check (5), exactly (5), onboarding (5), who (5), leadership (4), power (4), learning (4), analysis (4), python (4), make (4), window (4), discover (4), world (4), complex (4), arrow (4), using (4), artificial (4), topics (4), effort (4), hard (4), images (4), sessions (4), improvements (4), million (4), project (4), cybersecurity (4), safeguards (4), less (4), capable (4), research (4), being (4), broader (4), rather (4), especially (4), like (4), ceiling (4), most (4), point (4), session (4), other (4), because (4), full (4), comparison (4), task (4), these (4), through (4), three (4), out (4), been (4), worth (4), earlier (4), verified (4), slightly (4), without (4), areas (4), current (4), even (4), flagship (4), 280 (4), image (4), had (4), low (4), include (4), document (4), state (4), list (4), quoted (4), will (4), starts (4), 2026 (3), terms (3), security (3), information (3), stories (3), plan (3), plans (3), tutorials (3), azure (3), tableau (3), analyst (3), sql (3), scientist (3), visualization (3), cloud (3), career (3), skills (3), farnschläder (3), explore (3), josef (3), waples (3), tops (3), plus (3), start (3), details (3), build (3), applications (3), solutions (3), testing (3), cyber (3), broadly (3), while (3), says (3), rates (3), public (3), interesting (3), way (3), means (3), people (3), short (3), runs (3), let (3), itself (3), budget (3), issues (3), alongside (3), conversation (3), question (3), compared (3), ahead (3), area (3), leads (3), mcp (3), atlas (3), openai (3), result (3), reported (3), multi (3), shows (3), competitor (3), development (3), questions (3), might (3), top (3), below (3), oland (3), axis (3), 130 (3), 220 (3), bar (3), numbers (3), 1780 (3), title (3), exports (3), imports (3), scotland (3), ireland (3), letter (3), could (3), job (3), commentary (3), two (3), multimodal (3), end (3), 320 (3), ctrl (3), fictional (3), sentences (3), historically (3), time (3), advisory (3), firm (3), senior (3), partners (3), notes (3), content (3), news (3), notice (2), linkedin (2), become (2), center (2), français (2), deutsch (2), português (2), español (2), learner (2), program (2), teams (2), donates (2), book (2), demo (2), open (2), source (2), docs (2), alongs (2), fundamentals (2), associate (2), certification (2), datalab (2), statistics (2), machine (2), excel (2), sheets (2), aws (2), alteryx (2), tracks (2), mobile (2), capabilities (2), finance (2), khalid (2), abdelaty (2), structure (2), workflow (2), yet (2), extract (2), scope (2), write (2), ever (2), including (2), github (2), editor (2), tier (2), gives (2), latency (2), default (2), literal (2), input (2), much (2), glasswing (2), week (2), under (2), broad (2), capability (2), keeping (2), those (2), trained (2), showing (2), technical (2), writes (2), author (2), scientific (2), amount (2), makes (2), clear (2), going (2), actually (2), longer (2), running (2), sharper (2), were (2), usage (2), raised (2), rate (2), account (2), some (2), literally (2), need (2), tune (2), prompts (2), cheaper (2), mid (2), closely (2), them (2), autonomy (2), general (2), budgets (2), run (2), permissions (2), setting (2), decisions (2), step (2), entirely (2), auto (2), mode (2), dedicated (2), pass (2), would (2), ultrareview (2), sits (2), depth (2), updates (2), gets (2), measured (2), products (2), practical (2), financial (2), mention (2), definitely (2), osworld (2), anyone (2), production (2), ability (2), complete (2), muse (2), spark (2), very (2), accuracy (2), largest (2), gain (2), major (2), gpqa (2), diamond (2), cannot (2), graduate (2), pretty (2), essentially (2), designed (2), difficult (2), variant (2), lead (2), last (2), though (2), released (2), terminal (2), competitors (2), table (2), nland (2), although (2), rubric (2), knew (2), since (2), playfair (2), 110 (2), then (2), tens (2), 150 (2), 170 (2), 200 (2), 240 (2), 260 (2), correctly (2), type (2), uncertain (2), years (2), 1781 (2), ten (2), thousand (2), pounds (2), christmas (2), places (2), duplicate (2), footnote (2), lines (2), layout (2), single (2), hand (2), place (2), name (2), subject (2), trade (2), mark (2), fed (2), degraded (2), article (2), may (2), possible (2), preamble (2), explanation (2), describe (2), medium (2), number (2), believe (2), refers (2), character (2), distinct (2), contain (2), sentence (2), second (2), applied (2), went (2), limit (2), into (2), asterisks (2), named (2), once (2), headings (2), british (2), spelling (2), separated (2), blank (2), surname (2), quotes (2), material (2), paragraphs (2), share (2), engagement (2), understanding (2), organisation (2), yourself (2), meet (2), writing (2), own (2), english (2), made (2), did (2), file (2), reads (2), follows (2), such (2), large (2), language (2), team (2), platform (2), training (2), product (2), roles (2), blogs (2), inc, rights, reserved, accessibility, sell, personal, cookie, privacy, policy, instagram, youtube, twitter, facebook, affiliate, help, contact, press, instructor, careers, customer, unlimited, expense, discounts, promos, sales, universities, students, portfolio, rdocumentation, upcoming, events, resource, resources, certified, engineer, certifications, documentation, started, probability, roadmap, skill, assessments, progress, daily, minute, challenges, grow, featuring, near, advanced, analyzing, guide, selection, compares, derrick, mwiti, find, matt, crabtree, leaderboards, related, solve, powered, introduction, value, opportunities, create, pocs, implement, develop, strategy, implementing, faster, developer, copilot, windsurf, replit, track, sports, analytics, enthusiast, finer, control, versus, cost, initiative, announced, focused, risks, benefits, framework, decided, hold, automatic, detection, blocking, risk, requests, deployment, professionals, legitimate, needs, vulnerability, penetration, red, teaming, apply, verification, uncapped, limited, trials, vehicles, aligned, date, lowest, misaligned, behavior, evaluations, timeline, staged, suggests, treating, possibility, faqs, educator, manages, posts, previously, worked, deutsche, telekom, writer, contributions, articles, journals, interested, linear, algebra, play, fair, chess, framing, constrained, vehicle, class, shipping, handle, releases, forward, direction, supervised, cross, backing, conclusion, concerned, part, limits, totally, true, offset, reporting, pipelines, acting, warned, saying, breakdown, scenarios, pick, post, 272k, reprices, flat, browsecomp, browser, response, steering, desktop, dense, looked, contained, directly, basic, takeaway, optimizes, horizon, unified, purpose, developers, cap, spend, useful, want, pace, burn, behalf, middle, ground, approving, skipping, gained, slash, command, review, changes, flags, bugs, design, careful, human, reviewer, catch, free, ultrareviews, try, extra, another, option, choosing, recommends, starting, shipped, knowing, least, else, launched, charge, priciest, powerful, rough, ratio, costs, reminder, message, send, includes, whole, far, forth, reply, whatever, refills, rolling, always, worries, inflation, thankfully, price, increase, microsoft, foundry, vertex, amazon, bedrock, making, wait, oriented, vending, money, too, average, ended, 937, 018, deserve, honorable, leaderboard, significantly, excels, meaningful, margin, competition, flagships, strong, building, systems, rely, orchestration, controlling, interface, clicking, typing, navigating, just, highest, likely, reasons, screenspot, resolutions, immense, boost, relative, say, field, substantial, charxiv, measures, charts, figures, mirror, focus, targets, biology, fields, tested, specifically, jump, noting, usually, saturated, days, four, achieve, tied, collection, mathematics, humanities, raw, external, retrieval, gap, favor, clearest, humanity, exam, said, compete, regard, self, caveat, mind, comparing, disappointing, newest, doesn, central, quite, improves, predecessor, overtakes, headroom, resolve, repositories, harder, percentage, wise, puts, clearly, reaches, evaluated, covering, thing, struck, funny, know, show, uncertainty, really, confidently, poland, greenland, didn, official, fairly, give, tricky, genuinely, switch, counting, partway, odd, weird, horizontal, antique, engraving, ticks, 300, 000, division, parts, year, header, names, jersey, isle, man, prussia, portugal, holland, sweden, guernsey, germany, denmark, norway, flanders, west, indies, america, russia, upright, divisions, black, ribbed, panel, plot, paired, bars, column, couldn, corner, exact, values, whether, bottom, reading, saw, version, wikipedia, validate, examples, screenshot, 1786, pulled, analysing, blurry, otherwise, accurate, requested, deliverable, visualisation, appears, dashboard, report, page, however, partial, ambiguous, readings, 138, 188, possibilities, piece, text, titles, labels, legends, footnotes, precisely, unclear, enue, many, sections, panels, visible, appear, overall, story, flag, anything, attempted, got, failure, consider, 400, doing, perfect, 325, save, trouble, cored, paste, counter, italicised, bullets, organization, color, realize, etc, ends, attributions, quotation, marks, attributed, banned, look, breaks, constraint, welcome, harding, whitcombe, treated, foundation, successful, term, centres, behind, begins, weeks, devoted, almost, listening, clients, assigned, coordinates, introductions, tax, audit, ensures, colleague, touching, shares, consistent, priorities, sensitivities, operated, transactional, handled, piecemeal, whichever, signed, changed, 2014, recognised, winning, relationships, fees, alone, helena, marchbanks, managing, often, reminds, hires, feels, understood, rarely, invoice, month, six, revised, measurably, retention, referral, service, expected, attend, initial, discovery, meeting, purely, observational, capacity, familiarise, sector, recent, filings, onboarded, thornbury, mills, ltd, spring, junior, annual, reports, spotted, inventory, anomaly, shaped, entire, brief, preparation, standard, expect, exception, emphasise, candour, outset, particularly, around, realistic, timelines, lose, pitch, promise, something, deliver, rupert, castellane, head, practice, philosophy, colours, take, ask, observe, action, recognise, trust, accumulates, slowly, pays, compounding, dividends, ready, employee, sized, accounting, follow, requirements, begin, fewer, streamline, leverage, robust, synergy, anywhere, direct, attribute, quote, fourth, toward, requirement, throughout, colour, realise, bullet, points, numbered, lists, specific, italicise, markdown, total, after, used, detailed, prompt, stress, constraints, created, poke, claims, tells, significant, upshot, keep, think, kept, good, judgment, down, reliable, recall, later, turns, lets, front, explained, conventions, fifth, appreciate, root, uses, obvious, system, based, pattern, files, works, future, megapixels, times, older, accepted, extraction, profit, heavily, increased, granularity, detail, they, depend, readability, sometimes, pixel, helpful, feature, kind, depends, reliability, unintended, consequences, urges, harnesses, skipped, interpreted, loosely, taken, seen, notable, mentioned, faq, sibling, deliberately, reduced, scaled, focuses, demanding, require, sustained, sitting, family, haiku, walk, dig, guides, recently, meta, held, isn, researcher, iteration, headline, pitching, hardest, longest, close, supervision, bespoke, solution, group, updated, apr, locked, away, home, browse, thought, mlops, deep, transformation, storytelling, literacy, governance, services, big, sqlite, spreadsheets, snowflake, sigma, scala, pyspark, postgresql, nosql, mysql, mongodb, kubernetes, kafka, julia, java, hadoop, git, generative, flink, docker, dbt, databricks, chatgpt, airflow, technology, technologies, leader, contribute, enterprise, industry, leaders, industries, life, classrooms, newsletter, cheat, podcasts, found, error, ไทย, svenska, русский, română, polski, 한국어, 日本語, हिन्दी, nederlands, tiếng, việt, bahasa, indonesia, türkçe, italiano, skip, main,

Text of the page (random words):
ing detailed prompt to stress test it it has 12 distinct constraints which we can check at the end you are writing a short briefing document for a new employee at a mid sized accounting firm follow every instruction below write a briefing about the firm s approach to client onboarding the briefing must meet all of the following requirements the briefing must be exactly 4 paragraphs long the second paragraph must begin with the word historically each paragraph must contain between 3 and 5 sentences no paragraph may have fewer than 3 or more than 5 do not use the words streamline leverage robust or synergy anywhere in the document include exactly two direct quotes from fictional senior partners attribute each quote to a named partner do not include any other quoted material one of the partners quoted must have a surname that starts with the letter m end the document with a single sentence closing line on its own separated from the fourth paragraph by a blank line this closing line does not count toward the 4 paragraph requirement the closing line must be a question use british english spelling throughout e g organisation colour realise do not use any bullet points numbered lists or headings mention a specific fictional client name exactly once and italicise it using markdown asterisks the total word count of the briefing including the closing line must be between 280 and 320 words do not include any preamble explanation or commentary before or after the briefing output only the briefing itself and here is the result welcome to harding whitcombe where client onboarding is treated as the foundation of every successful long term engagement our approach centres on understanding the organisation behind the numbers before any technical work begins which means the first two weeks are devoted almost entirely to listening new clients are assigned a dedicated onboarding lead who coordinates introductions across tax audit and advisory teams this structure ensures that every colleague touching the account shares a consistent understanding of the client s priorities and sensitivities historically the firm operated a more transactional model with onboarding handled piecemeal by whichever partner had signed the engagement letter that changed in 2014 when senior leadership recognised that our competitors were winning work on relationships rather than fees alone as helena marchbanks our managing partner often reminds new hires a client who feels understood in week one rarely questions the invoice in month six the revised approach has measurably improved retention and referral rates across every service line in practical terms you will be expected to attend the initial discovery meeting for any client you support even in a purely observational capacity you should familiarise yourself with the client s sector recent filings and any public commentary from their leadership before that first conversation when we onboarded thornbury mills ltd last spring the junior associate who had read three years of their annual reports was the one who spotted the inventory anomaly that shaped our entire advisory brief preparation of that depth is the standard we expect not the exception you will also notice that we emphasise candour from the outset particularly around scope and realistic timelines we would rather lose a pitch than promise something we cannot deliver says rupert castellane head of our advisory practice and that philosophy colours every onboarding conversation take the time to ask questions observe the senior partners in action and recognise that trust accumulates slowly but pays compounding dividends are you ready to meet your first client now is time to score it i ll share our rubric constraint how to check 1 exactly 4 paragraphs plus closing line count paragraph breaks 2 paragraph 2 starts with historically look at first word 3 every paragraph has 3 5 sentences count sentences per paragraph 4 no banned words ctrl f for each 5 exactly 2 quotes both attributed no other quoted material count quotation marks 6 one partner s surname starts with m check attributions 7 closing line separated by blank line visual check 8 closing line is a question ends in 9 british spelling ctrl f for organization color realize etc 10 no bullets numbers headings visual check 11 fictional client named once italicised ctrl f for asterisks 12 word count 280 320 paste into word counter we will save you the trouble of going through it the model s cored 11 12 it had been doing a perfect job but in the end it went over the word count very slightly 325 words and a 320 word limit we applied the same test to claude sonnet 4 6 and it also got 11 12 of the instructions correctly with a failure on the same place word count but we still consider opus 4 7 to be an improvement because sonnet 4 6 went way way over the limit 400 words multimodal for our second test we fed opus 4 7 a hard to read image you are analysing a degraded image that may be blurry low resolution or otherwise difficult to read your job is to extract as much accurate information as possible complete all of the following tasks do not include any preamble explanation or commentary output only the requested deliverable 1 describe what type of document or visualisation this appears to be e g bar chart dashboard table report page state your confidence as high medium or low 2 list every number you can make out however partial for each one state what you believe it refers to and your confidence high medium or low if a number is ambiguous between two readings e g could be 138 or 188 state both possibilities 3 list every piece of text you can make out titles labels legends footnotes quoted as precisely as possible where a word or character is unclear use to mark the uncertain character e g re enue 4 describe the layout how many distinct sections or panels are visible and what does each appear to contain 5 state what you believe the overall subject or story of the image is in one sentence of no more than 25 words flag if your confidence is low 6 list anything you attempted to read but could not make out at all alongside these instructions we fed opus 4 7 a degraded screenshot of playfair s 1786 scotland trade chart which we pulled from our article on data visualization examples we knew we could validate the output because we had a higher resolution version from wikipedia here is the image that opus 4 7 saw here is opus 4 7 s reading type horizontal bar chart antique engraving confidence high numbers x axis ticks 10 20 30 110 130 150 170 200 220 240 260 280 l 300 000 high but 130 220 uncertain years 1780 and 1781 in title high ten thousand pounds per division high title exports and imports of scotland to and from different parts for one year from christmas 1780 to christmas 1781 header names of places places jersey c ireland oland isle of man gr oe nland prussia portugal holland sweden guernsey germany denmark and norway flanders west indies america russia ireland duplicate footnote the upright divisions are ten thousand pounds each the black lines are exports the ribbed lines imports layout single panel title plot area with 16 paired bars right hand place name column footnote below subject a 1780 81 trade chart showing scotland s exports and imports by partner confidence high couldn t read top right corner mark exact bar values first letter of oland whether bottom ireland is a duplicate we knew this test might be a tricky one since the x axis of playfair s chart genuinely does switch counting partway through every 10 up to 110 then odd tens 130 150 170 then even tens 200 220 240 260 280 but opus 4 7 read the axis correctly on the first pass even though it was weird one thing that still struck us as funny i know that we like when ai models show their uncertainty but oland really should have been read confidently as poland and gr oe nland should have been greenland although we didn t have an official rubric for this one we can fairly give it 9 10 claude opus 4 7 benchmarks opus 4 7 was evaluated across 14 benchmarks covering coding reasoning tool use computer use and visual reasoning the table below shows the full comparison with opus 4 6 gpt 5 4 gemini 3 1 pro and the not yet published mythos preview agentic coding on swe bench pro opus 4 7 scores 64 3 ahead of gpt 5 4 at 57 7 gemini 3 1 pro at 54 2 and opus 4 6 at 53 4 on swe bench verified opus 4 7 reaches 87 6 against gemini 3 1 pro s 80 6 and opus 4 6 s 80 8 gpt 5 4 has no published score on swe bench verified in this comparison swe bench tests a model s ability to resolve real github issues in open source python repositories pro is a harder variant with more complex issues the 10 9 point gain over opus 4 6 on swe bench pro is the largest improvement in this release percentage point wise and puts opus 4 7 clearly ahead of both major competitors on this task only anthropic s internal mythos preview scores 77 8 on swe bench pro and 93 9 on swe bench verified which shows there is still headroom above opus 4 7 still mythos is not broadly available so for production use opus 4 7 is the current ceiling the results in terms of terminal bench 2 0 which is the go to benchmark for agentic coding in the terminal are quite interesting opus 4 7 improves slightly against its predecessor 69 4 vs 65 4 and overtakes gemini 3 1 pro which scores 68 5 that being said it can t compete with gpt 5 4 75 1 in this regard even though the self reported caveat on gpt 5 4 s score is worth keeping in mind when comparing still it might be disappointing for anthropic that its newest released flagship model doesn t top a benchmark that is central to agentic software development reasoning humanity s last exam is a collection of graduate level questions across science mathematics and humanities designed to be difficult for current frontier models opus 4 7 leads the no tools variant which tests raw reasoning without external retrieval the with tools gap in favor of gpt 5 4 58 7 is the clearest area where opus 4 7 54 7 does not lead graduate level reasoning is usually measured in gpqa diamond which is pretty saturated these days like all current frontier models all four compared models achieve a score over 90 and are essentially tied a more interesting benchmark result targets biology one of the fields also tested in the gpqa diamond specifically while we cannot compare it to the competitor models the jump between opus 4 6 30 9 and opus 4 7 74 0 is definitely worth noting visual reasoning the results in the charxiv benchmark which measures visual reasoning about scientific charts and figures mirror that this was one of the major focus areas in the development of opus 4 7 the 13 point no tools improvement is the largest relative gain in this release without competitor scores to compare against it is hard to say where opus 4 7 sits in the broader field on this task but the improvement over opus 4 6 is substantial the support for higher resolution images is likely one of the reasons for the improved visual reasoning the chart of screenspot pro scores at different resolutions shows the immense accuracy boost opus 4 7 gets from a higher resolution especially without tool use 79 5 vs 69 0 agentic tool and computer use mcp atlas tests performance across complex multi tool workflows opus 4 7 scores 77 3 here which is the highest of any model in the comparison the only model that tops opus 4 7 s mcp atlas score very slightly is muse spark which is reported at 78 3 osworld tests a model s ability to complete tasks by controlling a computer interface clicking typing and navigating applications opus 4 7 scores 78 0 up from 72 7 in opus 4 6 gpt 5 4 scores 75 0 and gemini 3 1 pro has no published score on this benchmark mythos preview scores 79 6 just slightly above opus 4 7 this is definitely one area where opus 4 7 excels it leads both mcp atlas and osworld verified by a meaningful margin over the competition flagships from google and openai which is a strong result for anyone building production agent systems that rely on tool orchestration and computer use agents financial analysis the model s financial analysis skills deserve an honorable mention opus 4 7 tops the finance agent v1 1 leaderboard 64 4 being significantly ahead of both gpt 5 4 61 5 and gemini 3 1 pro 59 7 in the more practical oriented vending bench 2 we can see that opus 4 7 is better with money too on average the model ended up with an amount of 10 937 compared to 8 018 with opus 4 6 how to access claude opus 4 7 anthropic is not making us wait opus 4 7 is now available in all claude products the claude api claude opus 4 7 amazon bedrock vertex ai and microsoft foundry what about claude opus 4 7 pricing one always worries especially now about inflation but thankfully pricing is the same as opus 4 6 that s 5 per million input tokens and 25 per million output tokens there was also no price increase between opus 4 6 and opus 4 7 as a general reminder if you are using opus in claude ai every message you send includes the whole conversation so far your new question all earlier back and forth plus the model s reply all of that gets measured in tokens your plan whatever it is has a token budget that refills on a rolling window different models charge different rates against that budget opus is the priciest because it s the most powerful a rough published ratio opus costs about 5 what sonnet does per token what else launched alongside opus 4 7 anthropic shipped several updates alongside the model itself that are worth knowing at least if you use claude code or build on the api new xhigh effort level a new xhigh extra high effort level now sits between high and max this gives you another option when choosing between reasoning depth and latency in claude code the default effort level has been raised to xhigh across all plans and anthropic recommends starting with high or xhigh when testing opus 4 7 on coding and agentic tasks ultrareview in claude code claude code gained a new ultrareview slash command which runs a dedicated review pass over your changes and flags bugs or design issues that a careful human reviewer would catch pro and max users get three free ultrareviews to try it out auto mode for max users max users now get access to auto mode a permissions setting where claude makes decisions on your behalf it s a middle ground between approving every step and skipping permissions entirely task budgets on the api task budgets are now in public beta on the claude api these let developers cap claude s token spend across a run which is especially useful for longer agent workflows where you want the model to pace itself rather than burn through its budget claude opus 4 7 vs gpt 5 4 we looked closely at the benchmark results and those results contained in them a comparison to other models at a high level let s compare opus 4 7 with gpt 5 4 directly the basic takeaway is that opus 4 7 optimizes for long horizon autonomy gpt 5 4 is a unified general ...